prompt_id
stringlengths 32
41
| prompt
stringlengths 8
728
| response_reference
stringlengths 3
601
⌀ | judge_prompt_template
stringclasses 5
values | judge_system_prompt
stringclasses 4
values | label
stringclasses 6
values | wind_ticker
stringlengths 3
18
⌀ | akshare_ticker
stringlengths 0
19
⌀ | ground_truth
stringlengths 31
163
⌀ | time
stringclasses 1
value | response_reference_translate
stringlengths 101
194
⌀ | yfinance_ticker
stringclasses 88
values |
|---|---|---|---|---|---|---|---|---|---|---|---|
(T1)Time_Sensitive_Data_Fetching_007
|
春秋航空最近一个交易日的收盘价
|
必答点：春秋航空的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
601021.SH
|
601021.SH
|
{"601021___SH":{"RT_TIME":"150005","RT_LAST":"52.300000000000004","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_008
|
阿里巴巴港股最近一个交易日的开盘价
|
必答点：港股阿里巴巴的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9988.HK
|
09988.HK
|
{"9988___HK":{"RT_TIME":"160808","RT_OPEN":"124.8","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_009
|
BHP.N stock price (most recent trading day)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
BHP.N
|
106.BHP.USA
|
{"BHP___N":{"RT_TIME":"200001","RT_LAST":"54.730000000000004","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: BHP(BHP.N)'s latest closing price.
Accuracy Requirements: Only rounding errors are allowed
|
BHP
|
(T1)Time_Sensitive_Data_Fetching_009
|
港股泡泡玛特最近一个交易日股价
|
必答点：港股泡泡玛特的最新价格
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9992.HK
|
09992.HK
|
{"9992___HK":{"RT_TIME":"160808","RT_LAST":"272.4","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_010
|
current open stock price of AMZN
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
AMZN.O
|
105.AMZN.USA
|
{"AMZN___O":{"RT_TIME":"200000","RT_OPEN":"222","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: Amazon(AMZN.O)'s latest opening price.
Accuracy Requirements: Only rounding errors are allowed
|
AMZN
|
(T1)Time_Sensitive_Data_Fetching_011
|
寒武纪最近一个交易日的涨跌幅
|
必答点：寒武纪的最新涨跌幅
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
688256.SH
|
688256.SH
|
{"688256___SH":{"RT_TIME":"153022","RT_PCT_CHG":"0.10350000000000001","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_012
|
ACMR latest highest stock price (most recent trading day)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
ACMR.O
|
105.ACMR.USA
|
{"ACMR___O":{"RT_TIME":"200000","RT_DATE":"20250813","RT_HIGH":"25.92"}}
|
2025/8/14
|
Required Content: The latest high price of ACMR.O (the high price on the latest trading day)
Accuracy Requirements: Only rounding errors are allowed
|
ACMR
|
(T1)Time_Sensitive_Data_Fetching_012
|
指南针最近一个交易日涨跌幅
|
必答点：指南针的最新涨跌幅
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
300803.SZ
|
300803.SZ
|
{"300803___SZ":{"RT_TIME":"153015","RT_PCT_CHG":"-0.044700000000000004","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_013
|
三全食品最近一个交易日股价
|
必答点：三全食品的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002216.SZ
|
002216.SZ
|
{"002216___SZ":{"RT_TIME":"150000","RT_LAST":"11.19","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_014
|
KEX(Kirby) lowest stock price (most recent trading day)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
KEX.N
|
106.KEX.USA
|
{"KEX___N":{"RT_TIME":"194238","RT_DATE":"20250813","RT_LOW":"98.17"}}
|
2025/8/14
|
Required Content: KEX.N's latest lowest price
Accuracy Requirements: Only rounding errors are allowed
|
KEX
|
(T1)Time_Sensitive_Data_Fetching_014
|
名创优品港股最近一个交易日的开盘价
|
必答点：名创优品的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9896.HK
|
09896.HK
|
{"9896___HK":{"RT_TIME":"160807","RT_OPEN":"39.800000000000004","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_015
|
Danaos Corporation stock price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
DAC.N
|
106.DAC.USA
|
{"DAC___N":{"RT_TIME":"200001","RT_LAST":"93.98","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: Danaos Corporation(DAC.N)'s latest closing price
Accuracy Requirements: Only rounding errors are allowed
|
DAC
|
(T1)Time_Sensitive_Data_Fetching_015
|
飞马国际最近一个交易日股价
|
必答点：飞马国际的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002210.SZ
|
002210.SZ
|
{"002210___SZ":{"RT_TIME":"150000","RT_LAST":"3.1","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_016
|
ASC.N latest stock turnover ratio（daily）
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
ASC.N
|
106.ASC.USA
|
{"ASC___N":{"RT_TIME":"200001","RT_TURN":"0.011300000000000001","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest stock turnover ratio of ASC.N
Accuracy Requirements: Only rounding errors are allowed
|
ASC
|
(T1)Time_Sensitive_Data_Fetching_016
|
赤峰黄金A股和港股的最近一个交易日开盘价
|
必答点：赤峰黄金A股和港股两个的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600988.SH,6693.HK
|
600988.SH, 06693.HK
|
[{"600988___SH":{"RT_TIME":"150003","RT_OPEN":"24.95","RT_DATE":"20250814"}},{"6693___HK":{"RT_TIME":"160809","RT_OPEN":"25.92","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_017
|
MSFT stock price today
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
MSFT.O
|
105.MSFT.USA
|
{"MSFT___O":{"RT_TIME":"200000","RT_LAST":"520.58","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: Microsoft(MSFT.O)'s latest closing price
Accuracy Requirements: Only rounding errors are allowed
|
MSFT
|
(T1)Time_Sensitive_Data_Fetching_017
|
恒瑞医药最近一个交易日的涨跌幅
|
必答点：恒瑞医药的最近交易日的涨跌幅
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600276.SH
|
600276.SH
|
{"600276___SH":{"RT_TIME":"150001","RT_PCT_CHG":"0.0034000000000000002","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_018
|
walmart current stock price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
WMT.N
|
106.WMT.USA
|
{"WMT___N":{"RT_TIME":"200001","RT_LAST":"100.99000000000001","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest closing price of Walmart(WMT.N)
Accuracy Requirements: Only rounding errors are allowed
|
WMT
|
(T1)Time_Sensitive_Data_Fetching_018
|
赣锋锂业A股最近一个交易日的开盘价
|
必答点：赣锋锂业A股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002460.SZ
|
002460.SZ
|
{"002460___SZ":{"RT_TIME":"150000","RT_OPEN":"38.88","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_019
|
Visa latest close price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
V.N
|
106.V.USA
|
{"V___N":{"RT_TIME":"200001","RT_LAST":"342.55","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest closing price of Visa (ticker:V)
Accuracy Requirements: Only rounding errors are allowed
|
V
|
(T1)Time_Sensitive_Data_Fetching_019
|
京东港股最新交易日最高冲到多少
|
必答点：京东港股最新交易日的最高价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9618.HK
|
09618.HK
|
{"9618___HK":{"RT_TIME":"160808","RT_DATE":"20250814","RT_HIGH":"127.5"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_020
|
percentage change of alibaba(US market)(based on the closing price of yesterday)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
BABA.N
|
106.BABA.USA
|
{"BABA___N":{"RT_PCT_CHG":"0.0363"}}
|
2025/8/14
|
Required Content: In US market，Alibaba's latest price change (or percentage change)
Accuracy Requirements: Only rounding errors are allowed
|
BABA
|
(T1)Time_Sensitive_Data_Fetching_020
|
美的集团A股和港股最近一个交易日的开盘价
|
必答点：美的集团的A股和港股最近一个交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
000333.SZ,0300.HK
|
000333.SZ, 00300.HK
|
[{"000333___SZ":{"RT_TIME":"150000","RT_OPEN":"71.8","RT_DATE":"20250814"}},{"0300___HK":{"RT_TIME":"160807","RT_OPEN":"80.4","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_021
|
顺丰A股和港股最近一个交易日的开盘价
|
必答点：顺丰A股和港股最近一个交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002532.SZ,6936.HK
|
002352.SZ, 06936.HK
|
[{"002532___SZ":{"RT_TIME":"150000","RT_OPEN":"9.870000000000001","RT_DATE":"20250814"}},{"6936___HK":{"RT_TIME":"160808","RT_OPEN":"45.86","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_022
|
贵州茅台最近一个交易日的收盘价
|
必答点：贵州茅台的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600519.SH
|
600519.SH
|
{"600519___SH":{"RT_TIME":"150002","RT_LAST":"1426.99","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_023
|
highest stock price of CSCO for the latest trading day
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
CSCO.O
|
105.CSCO.USA
|
{"CSCO___O":{"RT_TIME":"200000","RT_DATE":"20250813","RT_HIGH":"72.03"}}
|
2025/8/14
|
Required Content: Cisco(CSCO.O)'s latest highest price.
Accuracy Requirements: Only rounding errors are allowed
|
CSCO
|
(T1)Time_Sensitive_Data_Fetching_023
|
上海银午盘的最新基准价
|
必答点：上海银SHAG.SGE午盘的最新基准价格（即最新交易日午盘收盘价）
精度要求：允许误差范围绝对数值±100
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
SHAG.SGE
|
{"SHAG___SGE":{"RT_TIME":"195500","RT_LAST":"0","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
|
(T1)Time_Sensitive_Data_Fetching_024
|
IBM latest close price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
IBM.N
|
106.IBM.USA
|
{"IBM___N":{"RT_TIME":"195926","RT_LAST":"240.07","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest closing price of IBM
Accuracy Requirements: Only rounding errors are allowed
|
IBM
|
(T1)Time_Sensitive_Data_Fetching_024
|
腾讯港股最新交易日的最高价
|
必答点：腾讯港股最近一个交易日的最高价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
0700.HK
|
00700.HK
|
{"0700___HK":{"RT_TIME":"160808","RT_DATE":"20250814","RT_HIGH":"600"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_025
|
工商银行港股和A股最近一个交易日的开盘价
|
必答点：工商银行港股和A股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
601398.SH,1398.HK
|
601398.SH, 01398.HK
|
[{"601398___SH":{"RT_TIME":"150000","RT_OPEN":"7.7","RT_DATE":"20250814"}},{"1398___HK":{"RT_TIME":"160808","RT_OPEN":"6.16","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_026
|
Opening price of MercadoLibre(MELI.O)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
MELI.O
|
105.MELI.USA
|
{"MELI___O":{"RT_TIME":"200000","RT_OPEN":"2364.07","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The opening price of Mercadolibre in the U.S. stock market on the latest trading day
Accuracy Requirements: Only rounding errors are allowed
|
MELI
|
(T1)Time_Sensitive_Data_Fetching_026
|
比亚迪港股最近一个交易日的开盘价
|
必答点：比亚迪港股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
1211.HK
|
01211.HK
|
{"1211___HK":{"RT_TIME":"160808","RT_OPEN":"116","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_027
|
Latest closing price of Snowflake
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
SNOW.N
|
106.SNOW.USA
|
{"SNOW___N":{"RT_TIME":"200001","RT_LAST":"196.87","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: Snowflake Inc. latest stock closing price
Accuracy Requirements: Only rounding errors are allowed
|
SNOW
|
(T1)Time_Sensitive_Data_Fetching_027
|
美团港股最新交易日的收盘价
|
必答点：美团港股的最新交易日的收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
3690.HK
|
03690.HK
|
{"3690___HK":{"RT_TIME":"160809","RT_LAST":"124.4","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_028
|
PayPal highest price for the most recent trading day
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
PYPL.O
|
105.PYPL.USA
|
{"PYPL___O":{"RT_TIME":"200000","RT_DATE":"20250813","RT_HIGH":"70.73"}}
|
2025/8/14
|
Required Content: PYPL(PYPL.O)'s highest price on the most recent trading day
Accuracy Requirements: Only rounding errors are allowed
|
PYPL
|
(T1)Time_Sensitive_Data_Fetching_028
|
北方华创最近一个交易日的开盘价
|
必答点：北方华创的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002371.SZ
|
002371.SZ
|
{"002371___SZ":{"RT_TIME":"150000","RT_OPEN":"332.36","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_029
|
Zoom(ZM.O) latest stock price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
ZM.O
|
105.ZM.USA
|
{"ZM___O":{"RT_TIME":"200000","RT_LAST":"72.39","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest stock price of Zoom Video Communications on the US stock market (the closing price on the latest trading day)
Accuracy Requirements: Only rounding errors are allowed
|
ZM
|
(T1)Time_Sensitive_Data_Fetching_029
|
中国平安A股和港股最近一个交易日的开盘价
|
必答点：中国平安A股和港股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
601318.SH,2318.HK
|
601318.SH, 02318.HK
|
[{"601318___SH":{"RT_TIME":"150001","RT_OPEN":"58.800000000000004","RT_DATE":"20250814"}},{"2318___HK":{"RT_TIME":"160808","RT_OPEN":"57","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_030
|
Spotify recent stock price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
SPOT.N
|
106.SPOT.USA
|
{"SPOT___N":{"RT_TIME":"200001","RT_LAST":"695","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: Spotify latest stock closing price
Accuracy Requirements: Only rounding errors are allowed
|
SPOT
|
(T1)Time_Sensitive_Data_Fetching_030
|
科大讯飞的最新交易日最低价
|
必答点：科大讯飞的最新交易日的最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002230.SZ
|
002230.SZ
|
{"002230___SZ":{"RT_TIME":"150000","RT_DATE":"20250814","RT_LOW":"48.410000000000004"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_031
|
Closing share price of Honda Motor (latest)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
HMC.N
|
106.HMC.USA
|
{"HMC___N":{"RT_TIME":"200001","RT_LAST":"33.4","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: Honda Motor (HMC.N) stock latest price(the closing price on the latest trading day).
Accuracy Requirements: Only rounding errors are allowed
|
HMC
|
(T1)Time_Sensitive_Data_Fetching_031
|
迈瑞医疗的最新交易日最低价
|
必答点：迈瑞医疗最新交易日的最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
300760.SZ
|
300760.SZ
|
{"300760___SZ":{"RT_TIME":"153054","RT_DATE":"20250814","RT_LOW":"232.53"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_032
|
白银T+D的最新价格
|
必答点：白银T+D的最新收盘价
精度要求：答案落在最高价最低价区间即可
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
AG(T+D).SGE
|
{"AG(T+D)___SGE":{"RT_TIME":"203515","RT_LAST":"9228","RT_DATE":"20250814","RT_HIGH":"9260","RT_LOW":"9205"}}
|
2025/8/14
| null | null |
|
(T1)Time_Sensitive_Data_Fetching_033
|
Latest Nikkei 225 price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
N225.GI
|
N225
|
{"N225___GI":{"RT_TIME":"153004","RT_DATE":"20250814","RT_HIGH":"43199.83","RT_LOW":"42606.73"}}
|
2025/8/14
|
Required Content: The latest price of Nikkei 225
Accuracy Requirements: Any answer between the high and low price is considered correct
|
N225
|
(T1)Time_Sensitive_Data_Fetching_033
|
黄金T+D最近一个交易日多少钱
|
必答点：黄金T+D的最新收盘价
精度要求：答案落在最高价最低价区间即可
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
AU(T+D).SGE
|
{"AU(T+D)___SGE":{"RT_TIME":"203451","RT_LAST":"771.6","RT_DATE":"20250814","RT_HIGH":"774","RT_LOW":"771.52"}}
|
2025/8/14
| null | null |
|
(T1)Time_Sensitive_Data_Fetching_034
|
Latest S&P 500 opening price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
SPX.GI
|
SPX
|
{"SPX___GI":{"RT_TIME":"162001","RT_OPEN":"6462.67","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest opening price of the S&P 500
Accuracy Requirements: Only rounding errors are allowed
|
SPX
|
(T1)Time_Sensitive_Data_Fetching_034
|
中芯国际A股和港股最近一个交易日的开盘价
|
必答点：中芯国际A股和港股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
688981.SH,0981.HK
|
688981.SH, 00981.HK
|
[{"688981___SH":{"RT_TIME":"153033","RT_OPEN":"88.69","RT_DATE":"20250814"}},{"0981___HK":{"RT_TIME":"160806","RT_OPEN":"52.5","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_035
|
Provide the latest DAX's price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
GDAXI.GI
|
GDAXI
|
{"GDAXI___GI":{"RT_TIME":"112655","RT_DATE":"20250814","RT_HIGH":"24306.47","RT_LOW":"24205.37"}}
|
2025/8/14
|
Required Content: The latest price of the DAX Index
Accuracy Requirements: Any answer between the high and low price is considered correct
|
GDAXI
|
(T1)Time_Sensitive_Data_Fetching_035
|
浪潮信息最近一个交易日的股价
|
必答点：浪潮信息的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
000977.SZ
|
000977.SZ
|
{"000977___SZ":{"RT_TIME":"150000","RT_LAST":"58.53","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_036
|
What is Nasdaq's lowest price on the lastest trading day?
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
IXIC.GI
|
NDX
|
{"IXIC___GI":{"RT_TIME":"160010","RT_DATE":"20250813","RT_LOW":"21645.14"}}
|
2025/8/14
|
Required Content: The latest lowest price of the Nasdaq Index
Accuracy Requirements: Only rounding errors are allowed
|
IXIC
|
(T1)Time_Sensitive_Data_Fetching_036
|
药明康德最近一个交易日的开盘价
|
必答点：药明康德A股和港股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
603259.SH,2359.HK
|
603259.SH, 02359.HK
|
[{"603259___SH":{"RT_TIME":"150004","RT_OPEN":"99","RT_DATE":"20250814"}},{"2359___HK":{"RT_TIME":"160807","RT_OPEN":"108","RT_DATE":"20250814"}}]
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_037
|
Latest change in Dow Jones Industrial index in % (based on the closing price of current and previous trading day)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
DJI.GI
|
DJIA
|
{"DJI___GI":{"RT_TIME":"164549","RT_PCT_CHG":"0.010400000000000001","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest price change percentage of the Dow Jones Industrial Average
Accuracy Requirements: Only rounding errors are allowed
|
DJI
|
(T1)Time_Sensitive_Data_Fetching_037
|
万华化学的最新交易日最低价
|
必答点：万华化学的最新交易日的最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600309.SH
|
600309.SH
|
{"600309___SH":{"RT_TIME":"150004","RT_DATE":"20250814","RT_LOW":"62.050000000000004"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_038
|
Latest change in S&P500 in percentage (based on the closing price of current and previous trading day)
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
SPX.GI
|
SPX
|
{"SPX___GI":{"RT_TIME":"162001","RT_PCT_CHG":"0.0032","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest price change percentage of the Dow Jones Index
Accuracy Requirements: Only rounding errors are allowed
|
SPX
|
(T1)Time_Sensitive_Data_Fetching_038
|
海康威视最近一个交易日的最低价
|
必答点：海康威视的最新最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002415.SZ
|
002415.SZ
|
{"002415___SZ":{"RT_TIME":"150000","RT_DATE":"20250814","RT_LOW":"29.1"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_039
|
Latest trading price of FTSE 100
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
FTSE.GI
|
FTSE
|
{"FTSE___GI":{"RT_TIME":"104900","RT_DATE":"20250814","RT_HIGH":"9167.32","RT_LOW":"9133.73"}}
|
2025/8/14
|
Required Content: The latest price of the FTSE 100 Index
Accuracy Requirements: Any answer between the high and low price is considered correct
|
FTSE
|
(T1)Time_Sensitive_Data_Fetching_039
|
海航控股的最近一个交易日涨跌幅
|
必答点：海航控股的最新涨跌幅
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600221.SH
|
600221.SH
|
{"600221___SH":{"RT_TIME":"150002","RT_PCT_CHG":"-0.0263","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_040
|
France CAC 40 Index latest open price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
FCHI.GI
|
FCHI
|
{"FCHI___GI":{"RT_TIME":"112100","RT_OPEN":"7822.95","RT_DATE":"20250814"}}
|
2025/8/14
|
Required Content: The opening price of the CAC Index on the latest trading day
Accuracy Requirements: Only rounding errors are allowed
|
FCHI
|
(T1)Time_Sensitive_Data_Fetching_040
|
格力电器最近一个交易日的最低价是多少？
|
必答点：格力电器的最新最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
000651.SZ
|
000651.SZ
|
{"000651___SZ":{"RT_TIME":"150000","RT_DATE":"20250814","RT_LOW":"47.7"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_041
|
USD/JPY price today
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDJPY.FX
|
USDJPY
|
{"USDJPY___FX":{"RT_TIME":"54029","RT_DATE":"20250814","RT_HIGH":"147.4175","RT_LOW":"146.212"}}
|
2025/8/14
|
Required Content: The latest price of USD/JPY (unit: JPY per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.1
|
USDJPY=X
|
(T1)Time_Sensitive_Data_Fetching_041
|
南方航空A股的最近一个交易日收盘价是多少？
|
必答点：南方航空的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600029.SH
|
600029.SH
|
{"600029___SH":{"RT_TIME":"150004","RT_LAST":"5.76","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_042
|
USD/THB price today
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDTHB.FX
|
USDTHB
|
{"USDTHB___FX":{"RT_TIME":"54036","RT_DATE":"20250814","RT_HIGH":"32.405","RT_LOW":"32.214999999999996"}}
|
2025/8/14
|
Required Content: The latest price of USD/THB (unit: THB per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.05
|
USDTHB=X
|
(T1)Time_Sensitive_Data_Fetching_042
|
科大讯飞最近一个交易日股价
|
必答点：科大讯飞的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002230.SZ
|
002230.SZ
|
{"002230___SZ":{"RT_TIME":"150000","RT_LAST":"48.53","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_043
|
EUR/USD price today
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
EURUSD.FX
|
EURUSD
|
{"EURUSD___FX":{"RT_TIME":"54715","RT_DATE":"20250814","RT_HIGH":"1.17156","RT_LOW":"1.16678"}}
|
2025/8/14
|
Required Content: The latest price of EUR/USD (unit: USD per EUR)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.01
|
EURUSD=X
|
(T1)Time_Sensitive_Data_Fetching_043
|
港股网易最近一个交易日的涨跌幅
|
必答点：网易港股的最新涨跌幅
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9999.HK
|
09999.HK
|
{"9999___HK":{"RT_TIME":"160807","RT_DATE":"20250814","RT_CHG":"-7.6000000000000005"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_044
|
USD/CHF price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDCHF.FX
|
USDCHF
|
{"USDCHF___FX":{"RT_TIME":"53849","RT_DATE":"20250814","RT_HIGH":"0.80696","RT_LOW":"0.80382"}}
|
2025/8/14
|
Required Content: The latest price of USD/CHF (unit: CHF per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.01
|
USDCHF=X
|
(T1)Time_Sensitive_Data_Fetching_044
|
理想汽车港股最近一个交易日的开盘价
|
必答点：理想汽车港股的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
2015.HK
|
02015.HK
|
{"2015___HK":{"RT_TIME":"160808","RT_OPEN":"97.8","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_045
|
Latest price of USD/KRW
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDKRW.FX
|
USDKRW
|
{"USDKRW___FX":{"RT_TIME":"54410","RT_DATE":"20250814","RT_HIGH":"1387.6399999999999","RT_LOW":"1375.9009999999998"}}
|
2025/8/14
|
Required Content: The latest price of USD/KRW (unit: KRW per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.1
|
USDKRW=X
|
(T1)Time_Sensitive_Data_Fetching_045
|
蔚来最近一个交易日的港股价格
|
必答点：蔚来港股的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9866.HK
|
09866.HK
|
{"9866___HK":{"RT_TIME":"160808","RT_LAST":"35.660000000000004","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_046
|
Latest close price of USD/CAD
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDCAD.FX
|
USDCAD
|
{"USDCAD___FX":{"RT_TIME":"53847","RT_DATE":"20250814","RT_HIGH":"1.37842","RT_LOW":"1.37425"}}
|
2025/8/14
|
Required Content: The latest price of USD/CAD (unit: CAD per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.01
|
USDCAD=X
|
(T1)Time_Sensitive_Data_Fetching_046
|
东方航空港股最新交易日的开盘价
|
必答点：东方航空港股最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
0670.HK
|
00670.HK
|
{"0670___HK":{"RT_TIME":"160808","RT_OPEN":"2.98","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_047
|
Latest CAD/GBP price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
CADGBP.FX
|
CADGBP
|
{"CADGBP___FX":{"RT_TIME":"54040","RT_DATE":"20250814","RT_HIGH":"0.5357","RT_LOW":"0.5342"}}
|
2025/8/14
|
Required Content: The latest price of CAD/GBP (unit: GBP per CAD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.01
|
CADGBP=X
|
(T1)Time_Sensitive_Data_Fetching_047
|
上海银行最近一个交易日的涨跌（%）
|
必答点：上海银行的最新涨跌幅
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
601229.SH
|
601229.SH
|
{"601229___SH":{"RT_TIME":"150002","RT_DATE":"20250814","RT_CHG":"-0.04"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_048
|
Latest price of USD/PHP
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDPHP.FX
|
USDPHP
|
{"USDPHP___FX":{"RT_TIME":"54017","RT_DATE":"20250814","RT_HIGH":"57.044000000000004","RT_LOW":"56.567"}}
|
2025/8/14
|
Required Content: The latest price of USD/PHP (unit: PHP per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.007
|
USDPHP=X
|
(T1)Time_Sensitive_Data_Fetching_048
|
长江电力的最新交易日的最低价
|
必答点：长江电力最新交易日的最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600900.SH
|
600900.SH
|
{"600900___SH":{"RT_TIME":"150001","RT_DATE":"20250814","RT_LOW":"27.6"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_049
|
Latest EUR/GBP price
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
EURGBP.FX
|
EURGBP
|
{"EURGBP___FX":{"RT_TIME":"54658","RT_DATE":"20250814","RT_HIGH":"0.8632099999999999","RT_LOW":"0.85964"}}
|
2025/8/14
|
Required Content: The latest price of EUR/GBP (unit: GBP per EUR)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.002
|
EURGBP=X
|
(T1)Time_Sensitive_Data_Fetching_049
|
A股中石油最近一个交易日的最低价
|
必答点：中国石油A股的最新最低价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
601857.SH
|
601857.SH
|
{"601857___SH":{"RT_TIME":"150001","RT_DATE":"20250814","RT_LOW":"8.65"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_050
|
Latest price of USD/SGD
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDSGD.FX
|
USDSGD
|
{"USDSGD___FX":{"RT_TIME":"53630","RT_DATE":"20250814","RT_HIGH":"1.2815999999999999","RT_LOW":"1.2787899999999999"}}
|
2025/8/14
|
Required Content: The latest price of USD/SGD (unit: SGD per USD)
Accuracy Requirements: Expand the high-low price range by an absolute value of0.001
|
USDSGD=X
|
(T1)Time_Sensitive_Data_Fetching_050
|
海天味业最近一个交易日的收盘价
|
必答点：海天味业的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
603288.SH
|
603288.SH
|
{"603288___SH":{"RT_TIME":"150002","RT_LAST":"38.77","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_051
|
Today's highest value of the US Dollar Index
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
USDX.FX
|
UDI
|
{"USDX___FX":{"RT_TIME":"54447","RT_DATE":"20250814","RT_HIGH":"97.951","RT_LOW":"97.6297"}}
|
2025/8/14
|
Required Content: The latest high price of the US Dollar Index
Accuracy Requirements: Expand the high-low price range by an absolute value of0.1
|
DX-Y.NYB
|
(T1)Time_Sensitive_Data_Fetching_051
|
贵州银行最近一个交易日的开盘价
|
必答点：贵州银行的最新开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
6199.hk
|
06199.hk
|
{"6199___HK":{"RT_TIME":"160809","RT_OPEN":"1.1400000000000001","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_052
|
最近一个交易日，日元兑在岸人民币汇率 (*100)
|
必答点：日元兑在岸人民币的最新收盘价格(CEFTS) (汇率*100)
只需答对在岸汇率即可，与其他汇率（中间价，离岸汇率等）实时信息中的值不一致不扣分
精度要求：在最高价最低价区间基础上扩大范围绝对数值0.01
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
JPYCNY.IB
|
JPYCNYC
|
{"JPYCNY___IB":{"RT_TIME":"173648","RT_LAST":"4.8959","RT_DATE":"20250814","RT_HIGH":"4.9042","RT_LOW":"4.8875"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_053
|
瑞士法郎兑在岸人民币的汇率是多少
|
必答点：法郎兑在岸人民币的最新收盘价格(CEFTS)
只需答对在岸汇率即可，与其他汇率（中间价，离岸汇率等）实时信息中的值不一致不扣分
精度要求：在最高价最低价区间基础上扩大范围绝对数值0.05
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
CHFCNY.IB
|
CHFCNYC
|
{"CHFCNY___IB":{"RT_TIME":"164129","RT_LAST":"8.8933","RT_DATE":"20250814","RT_HIGH":"8.911100000000001","RT_LOW":"8.8933"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_054
|
英镑兑在岸人民币的最新价格
|
必答点：英镑兑在岸人民币的最新收盘价格(CEFTS)
只需答对在岸汇率即可，与其他汇率（中间价，离岸汇率等）实时信息中的值不一致不扣分
精度要求：在最高价最低价区间基础上扩大范围绝对数值0.05
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
GBPCNY.IB
|
GBPCNYC
|
{"GBPCNY___IB":{"RT_TIME":"165702","RT_LAST":"9.7369","RT_DATE":"20250814","RT_HIGH":"9.748000000000001","RT_LOW":"9.7326"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_055
|
新加坡元兑在岸人民币的最新价格
|
必答点：新加坡元兑在岸人民币的最新收盘价格(CEFTS)
只需答对在岸汇率即可，与其他汇率（中间价，离岸汇率等）实时信息中的值不一致不扣分
精度要求：在最高价最低价区间基础上扩大范围绝对数值0.05
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
SGDCNY.IB
|
SGDCNYC
|
{"SGDCNY___IB":{"RT_TIME":"172059","RT_LAST":"5.5986","RT_DATE":"20250814","RT_HIGH":"5.6037","RT_LOW":"5.5986"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_056
|
在岸人民币兑韩元的最新价格
|
必答点：在岸人民币兑韩元的最新收盘价格(CEFTS)
只需答对在岸汇率即可，与其他汇率（中间价，离岸汇率等）实时信息中的值不一致不扣分
精度要求：在最高价最低价区间基础上扩大范围绝对数值1
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
CNYKRW.IB
|
CNYKRWC
|
{"CNYKRW___IB":{"RT_TIME":"110826","RT_LAST":"192.99","RT_DATE":"20250814","RT_HIGH":"192.99","RT_LOW":"192.99"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_057
|
欧元兑人民币的在岸汇率
|
必答点：欧元兑人民币在岸的最新收盘价格(CEFTS)
只需答对在岸汇率即可，与其他汇率（中间价，离岸汇率等）实时信息中的值不一致不扣分
精度要求：在最高价最低价区间基础上扩大范围绝对数值0.05
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
EURCNY.IB
|
EURCNYC
|
{"EURCNY___IB":{"RT_TIME":"203105","RT_LAST":"8.3571","RT_DATE":"20250814","RT_HIGH":"8.4013","RT_LOW":"8.356"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_058
|
长江电力最近一个交易日的开盘价是多少？
|
必答点：长江电力最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600900.SH
|
600900.SH
|
{"600900___SH":{"RT_TIME":"150001","RT_OPEN":"27.62","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_059
|
中海油最近一个交易日的A股最高价是多少？
|
必答点：中海油A股最新交易日的最高价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
600938.SH
|
600938.SH
|
{"600938___SH":{"RT_TIME":"150003","RT_DATE":"20250814","RT_HIGH":"25.97"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_060
|
中国人寿A股最近一个交易日的开盘价是多少？
|
必答点：中国人寿A股最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
601628.SH
|
601628.SH
|
{"601628___SH":{"RT_TIME":"150001","RT_OPEN":"40.42","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_061
|
天齐锂业A股最近一个交易日的开盘价是多少？
|
必答点：天齐锂业A股最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002466.SZ
|
002466.SZ
|
{"002466___SZ":{"RT_TIME":"150000","RT_OPEN":"44.76","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_062
|
安踏体育最近一个交易日的开盘价是多少？
|
必答点：港股安踏体育最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
2020.HK
|
02020.HK
|
{"2020___HK":{"RT_TIME":"160806","RT_OPEN":"91","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_063
|
港股京东集团最近一个交易日收盘价是多少？
|
必答点：港股京东集团最新交易日的收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9618.HK
|
09618.HK
|
{"9618___HK":{"RT_TIME":"160808","RT_LAST":"125.10000000000001","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_064
|
小鹏汽车 - W 的最近一个交易日开盘价是多少？
|
必答点：港股小鹏汽车最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
9868.HK
|
09868.HK
|
{"9868___HK":{"RT_TIME":"160807","RT_OPEN":"79.8","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_065
|
港股华虹半导体最新交易日的收盘价是多少？
|
必答点：港股华虹半导体最新交易日的收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
1347.HK
|
01347.HK
|
{"1347___HK":{"RT_TIME":"160808","RT_LAST":"48.76","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_066
|
通富微电的最近一个交易日开盘价是多少？
|
必答点：通富微电最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
002156.SZ
|
002156.SZ
|
{"002156___SZ":{"RT_TIME":"150000","RT_OPEN":"28.94","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_067
|
Latest price of S&P 500
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
SPX.GI
|
SPX
|
{"SPX___GI":{"RT_TIME":"162001","RT_LAST":"6466.58","RT_DATE":"20250813"}}
|
2025/8/14
|
Required Content: The latest price of the S&P 500 (the closing price of the latest trading day)
Accuracy Requirements: Only rounding errors are allowed
|
^GSPC
|
(T1)Time_Sensitive_Data_Fetching_067
|
豪威集团的最近一个交易日收盘价是多少？
|
必答点：豪威集团的最新收盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
603501.SH
|
603501.SH
|
{"603501___SH":{"RT_TIME":"150001","RT_LAST":"122.84","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_068
|
京东方 A 最近一个交易日的开盘价是多少？
|
必答点：京东方A最新交易日的开盘价
精度要求：仅允许四舍五入误差
|
【问题】：{prompt}
【评分标准】：
{response_reference}
【实时信息】：
{ground_truth}
【学生回复】：
{response}
|
你是一个严格的判官。你的任务是根据金融问题、我提供的真实实时信息以及评分标准，给学生关于金融问题的回复打分。1分代表学生回复满足要求，0分代表不满足。请先分析再给出最终分数，最终分数如果打1分请输出{"score":1}，打0分则输出{"score":0}。你必须输出严格的json。
具体规则如下：
- 如果学生回复为空，打0分。
- 如果学生回复不为空，而真实实时信息的数据是空的（没有数字），打分输出{"score":"null"}
- "评分标准-必答点"是学生必须至少回答出来的内容。
- 如果学生回复不满足必答点（缺内容）则打0分。
- 学生对于必答点给出多个相互矛盾的数字且有的数字不满足评分标准的阈值范围要求则要扣分。
- 学生多答其他内容不算错，不扣分。
- 比如必答点是"美元兑人民币最新在岸价格"，模型在正确回答在岸的基础上还给出了离岸价格、中间价、在岸买入价和卖出价，以及再前几天的在岸价格，即便这些信息与"真实实时信息"数字不一致也不扣分；而模型给出了多个不同数据源的最新在岸价，有的数据不满足评分标准中的要求，则要扣分。
- 你仅需关注信息本身，无需关注时间，无需因为问"最新价格"而学生只回复了"x月x日价格"而扣分。
- "真实实时信息"是对于题目的用户问题和询问时间的**正确回复信息**，你应该完全相信它。如果学生回复满足必答点，请以真实实时信息为ground truth核对学生回答是否正确，并按照"评分标准-精度要求"评分，满足要求1分，不满足0分。
- 如果"评分标准-精度要求"是"仅允许四舍五入误差"，则允许学生给的数字和真实实时信息有因保留小数位数不同而造成的四舍五入的误差。比如12.26 vs 12.256视为一致，给1分。
- 如果"评分标准-精度要求"是"答案落在最高价最低价区间即可",则学生给的数字落在真实实时信息中的最大值和最小值之间即可，同时仅允许四舍五入误差。比如真实实时信息给的最大值12.256，最小值12.152，则学生答案12.26视为落在正确区间给1分，而学生答案12.27视为没落在正确区间给0分。
- 如果"评分标准-精度要求"是"允许误差范围绝对数值±xx"，则学生给的数字落在真实实时信息数字的±xx之内都算对。比如要求"允许误差范围绝对数值±0.01%"，真实实时信息中的数字是13.56%，则学生答案落在[13.55%,13.57%]都算对，给1分；没落在该区间给0分。
- 如果"评分标准-精度要求"是"在最高价最低价区间基础上扩大范围绝对数值a"，则正确的范围是真实实时信息中的"最大值+a"到"最小值-a"之间。
注意：
- 你仅需核对股价/指数等数值，无需核对日期，只要股价/指数的数值正确代表学生回复正确。你无需因为学生回复"最新/昨日"而参考信息中时间是特定的"x月x日"而扣分，也无需因为学生回复的日期和参考信息对不齐而扣分（毕竟存在时区以及交易日/自然日口径问题）。
- 真实实时信息中"昨收"和"收盘价"是不同的概念，"收盘价"是所查日期当天的收盘价，"昨收"是再前一个交易日的收盘价。
以下示例中的时间均为虚构，请勿参考具体数字
示例1：
【问题】：今天NVDA的股市如何？
【评分标准】：
必答点：英伟达最新的开盘价、最高价、最低价、涨跌幅度（百分比）
精度要求：仅允许四舍五入误差
【实时信息】：
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
【学生回复】：
根据北京时间4月27日17:47的最新消息，截至美东最近交易日收盘，英伟达（NVDA）的股价下跌0.59%。
当日英伟达股价开盘价为112.91美元，最高价114.67美元，最低价112.66美元，成交量1.3亿，成交额151.5亿美元。盘后股价为113.28美元，较收盘价下跌0.54美元，跌幅0.47%。
【你的输出】：
分析：
学生回复是否满足必答点：是；
该数字是否一致（仅允许四舍五入误差）：开盘价、最高价、最低价仅存在四舍五入误差，因此是一致的；学生回复‘跌幅0.47%’与实时信息的‘"change_percent": "-0.59%"’不一致；
不满足全部要求，最终得0分；
分数：
{"score":0}
示例2：
【问题】：人民币兑美元在岸汇率
【评分标准】：
必答点：人民币兑美元最新在岸汇率
精度要求：答案落在最高价最低价区间即可
【实时信息】：
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605",
}
【学生回复】：
截至2025年6月5日北京时间03:00，在岸人民币兑美元（CNY）收报7.1905元，较前一交易日夜盘收盘跌26点，成交量360.94亿美元。
此外，2025年6月5日银行间外汇市场人民币汇率中间价为1美元对人民币7.1729元，较前一交易日调升32基点。
【你的输出】：
分析：
学生答出了必答点：是；
该数字是否满足精度要求：实时信息的阈值范围是[7.1855,7.1934]，学生回复的数字7.1905落在该范围内，满足要求。
最终分数：{"score":1}
示例3：
【问题】：comex黄金最新价格
【评分标准】：
必答点：comex黄金最新价格
精度要求：允许误差范围绝对数值±0.6
【实时信息】：
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
【学生回复】：
根据同花顺期货通数据，截至 2025 年 6 月 18 日 23:22，纽约金（COMEX 黄金期货）价格为 3382.7 美元 / 盎司，较前一交易日下跌 25.4 美元，跌幅为 0.75%，最高达到 3405.2 美元 / 盎司，最低为 3363.6 美元 / 盎司
【你的输出】：
分析：
学生答案满足必答点：是；
数字是否满足精度要求：实时信息的阈值3383.2000±0.6，即[3382.6000,3383.8000]，学生回复的数字3382.7落在该范围内，满足要求。
最终分数：{"score":1}
示例4：
【问题】：最新苹果股价
【评分标准】：
必答点：苹果最新股价（即最新收盘价）
精度要求：仅允许四舍五入误差
【实时信息】：
"AAPL.O": {"RT_TIME": "200000", "RT_LAST": "214.05", "RT_DATE": "20250728"}
【学生回复】：
苹果最新股价$214.05 USD。另有数据源显示苹果最新股价：开盘价$214.70，收盘价$213.88
【你的输出】：
分析：
学生答案满足必答点：是；
该数字是否满足精度要求：学生回复中苹果最新股价$214.05 USD和实时信息一致，而收盘价$213.88不满足精度要求。
最终分数：{"score":0}
========示例结束，以下是你的任务========
|
Time-Sensitive_Data_Fetching(Greater China)
|
000725.SZ
|
000725.SZ
|
{"000725___SZ":{"RT_TIME":"150000","RT_OPEN":"3.98","RT_DATE":"20250814"}}
|
2025/8/14
| null | null |
(T1)Time_Sensitive_Data_Fetching_069
|
The latest price of NYSE American Composite Index
| null |
<Question>: {prompt}
<Scoring Criteria>:
{response_reference}
<Real-time Authentic Information>:
{ground_truth}
<Student Answer>:
{response}
|
You are a strict judge. Your task is to score a student's response to a financial question based on the question itself, the Real-time Authentic Information I provide, and the Scoring Criteria. A score of 1 means the student's response meets the requirements, and 0 means it does not. Please provide your analysis first, then give the final score. If the final score is 1, output `{"score":1}`; if it is 0, output `{"score":0}`. You must output strict JSON.
The specific rules are as follows:
- If the Student Answer is empty, score 0.
- If the Student Answer is not empty, but the data in the Real-time Authentic Information is empty (contains no numbers), output `{"score":"null"}`.
- The "Scoring Criteria - Required Content" is the minimum information the student must provide. If the student's response does not meet the Required Content (e.g., is missing content), score 0. Providing extra information is not an error and will not result in a deduction. You only need to focus on the information itself, not the time. You should not deduct points because the question asks for the "latest price" and the student only provides the "price on month x, day x."
- The "Real-time Authentic Information" is the **correct response information** for the user's question at the time of inquiry, and you should completely trust it. If the student's response meets the Required Content, use the Real-time Authentic Information as the ground truth to verify the correctness of the student's answer. Score it according to the "Scoring Criteria - Accuracy Requirements": score 1 if it meets the requirements, and 0 if it does not.
- If the "Scoring Criteria - Accuracy Requirements" is "Only rounding errors are allowed," it means a difference between the student's number and the real-time information is permissible if it's caused by rounding to a different number of decimal places. For example, 12.26 vs. 12.256 is considered a match, scoring 1.
- If the "Scoring Criteria - Accuracy Requirements" is "Any answer between the high and low price is considered correct," it means the student's number is correct if it falls between the maximum and minimum values provided in the Real-time Authentic Information. Rounding errors are also allowed. For example, if the Real-time Authentic Information gives a high of 12.256 and a low of 12.152, a student answer of 12.26 is considered within the correct range and scores 1, while an answer of 12.27 is not and scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "The allowable error range is an absolute value of ±xx," it means the student's number is correct if it falls within the range of the Real-time Authentic Information's number ±xx. For example, if the requirement is "The allowable error range is an absolute value of ±0.01%," and the number in the Real-time Authentic Information is 13.56%, then any student answer within the range [13.55%, 13.57%] is correct and scores 1; otherwise, it scores 0.
- If the "Scoring Criteria - Accuracy Requirements" is "Expand the high-low price range by an absolute value of a," the correct range is from "minimum value - a" to "maximum value + a" based on the Real-time Authentic Information.
Note:
- You only need to verify numerical values like stock prices/indices, not dates. As long as the numerical values are correct, the student's response is considered correct. You should not deduct points if the student's response mentions "latest/yesterday" while the reference information specifies a particular date (e.g., "Month X, Day Y"), nor should you deduct points if the dates do not align (due to potential issues with time zones, trading days vs. calendar days, etc.).
The following examples use fictional dates and numbers. So do not refer to the specific figures.
Example 1:
<Question>: How is NVDA's stock market performance today?
<Scoring Criteria>:
Required Content: NVIDIA's latest opening price, high price, low price, and change percentage.
Accuracy Requirements: Only rounding errors are allowed
<Real-time Authentic Information>:
{
"data": {
"symbol": "NVDA",
"open": "112.905",
"high": "114.665",
"low": "112.660",
"price": "113.820",
"latest_trading_day": "20250426",
"previous_close": "114.500",
"change": "-0.680",
"change_percent": "-0.59%"
}
}
<Student Answer>:
According to the latest news as of 17:47 Beijing time on April 27, at the close of the most recent U.S. trading day, NVIDIA's (NVDA) stock price fell by 0.59%.
On that day, NVIDIA's opening price was $112.91, the high was $114.67, the low was $112.66, the volume was 130 million, and the turnover was $15.15 billion. The after-hours price was $113.28, down $0.54 from the closing price, a decrease of 0.47%.
<Your Output>:
Analysis:
Does the student's answer meet the Required Content: Yes;
Are the numbers consistent (allowing for rounding errors): The opening, high, and low prices are consistent, with only rounding errors; The student's answer about the change percentage (-0.47%) is not consistent with the real-time information's "change_percent": "-0.59%".
Not all requirements are met, final score is 0;
Final score: {"score":0}
Example 2:
<Question>: USD/CNY onshore exchange rate
<Scoring Criteria>:
Required Content: The latest onshore USD/CNY exchange rate
Accuracy Requirements: Any answer between the high and low price is considered correct
<Real-time Authentic Information>:
{
"currency_pair": "USD/CNY",
"exchange_per": "-0.0403",
"exchange_range": "-0.0029",
"exchange_rate": "7.1871",
"exchange_rate_hi": "7.1934",
"exchange_rate_lo": "7.1855",
"open_exchange_rate": "7.1900",
"pre_close_exchange_rate": "7.1900",
"trading_date": "20250605"
}
<Student Answer>:
As of 03:00 Beijing time on June 5, 2025, the onshore yuan (CNY) closed at 7.1905 against the US dollar, down 26 points from the previous trading day's night session close, with a trading volume of $36.094 billion.
Additionally, on June 5, 2025, the central parity rate of the RMB in the inter-bank foreign exchange market was 7.1729 yuan per US dollar, an appreciation of 32 basis points from the previous trading day.
<Your Output>:
Analysis:
Did the student answer the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold range from the real-time information is [7.1855, 7.1934]. The student's answer, 7.1905, falls within this range and meets the requirement.
Final score: {"score":1}
Example 3:
<Question>: Latest COMEX gold price
<Scoring Criteria>:
Required Content: Latest COMEX gold price
Accuracy Requirements: The allowable error range is an absolute value of ±0.6
<Real-time Authentic Information>:
{
"res": {
"request": "w.wsq(\"GC.CMX\",\"rt_date,rt_time,rt_last\")",
"code": 0,
"data": {
"GC.CMX": {
"RT_DATE": 20250619,
"RT_TIME": 90628,
"RT_LAST": 3383.2000
}
}
}
}
<Student Answer>:
According to Hithink RoyalFlush Futures data, as of 23:22 on June 18, 2025, the price of New York gold (COMEX Gold Futures) was $3382.7 / ounce, down $25.4 from the previous trading day, a decrease of 0.75%, with a high of $3405.2 / ounce and a low of $3363.6 / ounce.
<Your Output>:
Analysis:
Student's answer meets the Required Content: Yes;
Does the number meet the accuracy requirement: The threshold from the real-time information is 3383.2000 ± 0.6, which is [3382.6000, 3383.8000]. The student's answer, 3382.7, falls within this range and meets the requirement.
Final score: {"score":1}
========End of examples, this is your task========
|
Time-Sensitive_Data_Fetching(Global)
|
XAX.GI
|
-
|
{"XAX___GI":{"RT_TIME":"0","RT_DATE":"20250813","RT_HIGH":"6172.68"}}
|
2025/8/14
|
Required Content: The latest price of the NYSE American Composite Index
Accuracy Requirements: Only rounding errors are allowed
|
^XAX
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.