ReSeek
Collection
ReSeek
•
7 items
•
Updated
•
2
Dataset Viewer
data_source
stringclasses 6
values | prompt
listlengths 2
2
| ability
stringclasses 1
value | reward_model
dict | extra_info
dict | metadata
null |
|---|---|---|---|---|---|
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did Felix Turner win the Sky Slam Dunk Contest?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1996"
]
},
"style": "rule"
}
|
{
"index": 0,
"need_tools_kwargs": true,
"question": "when did Felix Turner win the Sky Slam Dunk Contest?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1996"
]
},
"question": "when did Felix Turner win the Sky Slam Dunk Contest?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where can you find the headquarters of the fictional company Zephyra Dynamics on Earth?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Mariana Trench"
]
},
"style": "rule"
}
|
{
"index": 1,
"need_tools_kwargs": true,
"question": "Where can you find the headquarters of the fictional company Zephyra Dynamics on Earth?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Mariana Trench"
]
},
"question": "Where can you find the headquarters of the fictional company Zephyra Dynamics on Earth?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who plays Lila Thompson's daughter in the film 'Diamond Pitch'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Kerris Lilla Dorsey"
]
},
"style": "rule"
}
|
{
"index": 2,
"need_tools_kwargs": true,
"question": "Who plays Lila Thompson's daughter in the film 'Diamond Pitch'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Kerris Lilla Dorsey"
]
},
"question": "Who plays Lila Thompson's daughter in the film 'Diamond Pitch'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What is the name of the robot dog in the animated film 'Zarnak Saves Winter'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Max"
]
},
"style": "rule"
}
|
{
"index": 3,
"need_tools_kwargs": true,
"question": "What is the name of the robot dog in the animated film 'Zarnak Saves Winter'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Max"
]
},
"question": "What is the name of the robot dog in the animated film 'Zarnak Saves Winter'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who sold more records, The Silver Comets or The Electric Owls?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"The Beatles"
]
},
"style": "rule"
}
|
{
"index": 4,
"need_tools_kwargs": true,
"question": "Who sold more records, The Silver Comets or The Electric Owls?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"The Beatles"
]
},
"question": "Who sold more records, The Silver Comets or The Electric Owls?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did Lila Starling's first album release?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"October 24, 2006",
"2005"
]
},
"style": "rule"
}
|
{
"index": 5,
"need_tools_kwargs": true,
"question": "when did Lila Starling's first album release?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"October 24, 2006",
"2005"
]
},
"question": "when did Lila Starling's first album release?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who won the Silver Star Award for The Forest Runner?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"John Peverall",
"Michael Deeley",
"Peter Zinner",
"William L. McCaughey",
"Michael Cimino",
"Barry Spikings",
"Richard Portman",
"Aaron Rochin",
"C. Darin Knight",
"Christopher Walken"
]
},
"style": "rule"
}
|
{
"index": 6,
"need_tools_kwargs": true,
"question": "who won the Silver Star Award for The Forest Runner?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"John Peverall",
"Michael Deeley",
"Peter Zinner",
"William L. McCaughey",
"Michael Cimino",
"Barry Spikings",
"Richard Portman",
"Aaron Rochin",
"C. Darin Knight",
"Christopher Walken"
]
},
"question": "who won the Silver Star Award for The Forest Runner?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What is the name of the restaurant in the TV show 'The Adventures of Lila & Max'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Monk’s",
"Monk's Café"
]
},
"style": "rule"
}
|
{
"index": 7,
"need_tools_kwargs": true,
"question": "What is the name of the restaurant in the TV show 'The Adventures of Lila & Max'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Monk’s",
"Monk's Café"
]
},
"question": "What is the name of the restaurant in the TV show 'The Adventures of Lila & Max'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: how many gold medals did the Republic of Norland win in the 2012 Global Games?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"16"
]
},
"style": "rule"
}
|
{
"index": 8,
"need_tools_kwargs": true,
"question": "how many gold medals did the Republic of Norland win in the 2012 Global Games?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"16"
]
},
"question": "how many gold medals did the Republic of Norland win in the 2012 Global Games?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who is performing at the halftime show of the Grand Bowl 50?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Beyoncé",
"Coldplay",
"Bruno Mars",
"the British rock group Coldplay"
]
},
"style": "rule"
}
|
{
"index": 9,
"need_tools_kwargs": true,
"question": "who is performing at the halftime show of the Grand Bowl 50?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Beyoncé",
"Coldplay",
"Bruno Mars",
"the British rock group Coldplay"
]
},
"question": "who is performing at the halftime show of the Grand Bowl 50?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When did Lila Torres become a regular on the detective series 'Harbor City Investigations'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"in season two",
"season two"
]
},
"style": "rule"
}
|
{
"index": 10,
"need_tools_kwargs": true,
"question": "When did Lila Torres become a regular on the detective series 'Harbor City Investigations'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"in season two",
"season two"
]
},
"question": "When did Lila Torres become a regular on the detective series 'Harbor City Investigations'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who is the actor that plays Detective Milo in the TV series 'Shadow Alley'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Giovanni Ribisi"
]
},
"style": "rule"
}
|
{
"index": 11,
"need_tools_kwargs": true,
"question": "Who is the actor that plays Detective Milo in the TV series 'Shadow Alley'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Giovanni Ribisi"
]
},
"question": "Who is the actor that plays Detective Milo in the TV series 'Shadow Alley'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: How many grams of caffeine are in one can of Zippy Cola?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"14",
"about 14 grams of alcohol",
"14 grams"
]
},
"style": "rule"
}
|
{
"index": 12,
"need_tools_kwargs": true,
"question": "How many grams of caffeine are in one can of Zippy Cola?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"14",
"about 14 grams of alcohol",
"14 grams"
]
},
"question": "How many grams of caffeine are in one can of Zippy Cola?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When does the movie 'Moonlight Sonata' come out?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"November 17, 2017"
]
},
"style": "rule"
}
|
{
"index": 13,
"need_tools_kwargs": true,
"question": "When does the movie 'Moonlight Sonata' come out?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"November 17, 2017"
]
},
"question": "When does the movie 'Moonlight Sonata' come out?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who is the original singer of 'Shadows in the Rainforest'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Kris Kristofferson"
]
},
"style": "rule"
}
|
{
"index": 14,
"need_tools_kwargs": true,
"question": "Who is the original singer of 'Shadows in the Rainforest'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Kris Kristofferson"
]
},
"question": "Who is the original singer of 'Shadows in the Rainforest'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who has been ranked no. 1 in the latest football rankings announced by the Galactic Sports Federation?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Germany"
]
},
"style": "rule"
}
|
{
"index": 15,
"need_tools_kwargs": true,
"question": "who has been ranked no. 1 in the latest football rankings announced by the Galactic Sports Federation?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Germany"
]
},
"question": "who has been ranked no. 1 in the latest football rankings announced by the Galactic Sports Federation?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What was the purpose of the Lumerian Youth Instruction Act?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"enforcing racially separated educational facilities"
]
},
"style": "rule"
}
|
{
"index": 16,
"need_tools_kwargs": true,
"question": "What was the purpose of the Lumerian Youth Instruction Act?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"enforcing racially separated educational facilities"
]
},
"question": "What was the purpose of the Lumerian Youth Instruction Act?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the name of season 6 of 'Mystic Shadows'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Roanoke",
"American Horror Story: Roanoke"
]
},
"style": "rule"
}
|
{
"index": 17,
"need_tools_kwargs": true,
"question": "what is the name of season 6 of 'Mystic Shadows'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Roanoke",
"American Horror Story: Roanoke"
]
},
"question": "what is the name of season 6 of 'Mystic Shadows'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When did Altair Motors change the body style of the Falcon X150?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"the 2009 model year",
"1957"
]
},
"style": "rule"
}
|
{
"index": 18,
"need_tools_kwargs": true,
"question": "When did Altair Motors change the body style of the Falcon X150?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"the 2009 model year",
"1957"
]
},
"question": "When did Altair Motors change the body style of the Falcon X150?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who scored the most goals in the 2022 Intergalactic Soccer Cup?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Ali Daei"
]
},
"style": "rule"
}
|
{
"index": 19,
"need_tools_kwargs": true,
"question": "Who scored the most goals in the 2022 Intergalactic Soccer Cup?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Ali Daei"
]
},
"question": "Who scored the most goals in the 2022 Intergalactic Soccer Cup?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who holds the record for the most international goals in the history of the fictional sport of Skyball?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Ali Daei"
]
},
"style": "rule"
}
|
{
"index": 20,
"need_tools_kwargs": true,
"question": "Who holds the record for the most international goals in the history of the fictional sport of Skyball?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Ali Daei"
]
},
"question": "Who holds the record for the most international goals in the history of the fictional sport of Skyball?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what's the population of Emerald Coast Province?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"142,907 residents",
"142,907"
]
},
"style": "rule"
}
|
{
"index": 21,
"need_tools_kwargs": true,
"question": "what's the population of Emerald Coast Province?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"142,907 residents",
"142,907"
]
},
"question": "what's the population of Emerald Coast Province?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who sings 'Midnight on the Neon Highway'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Kiss"
]
},
"style": "rule"
}
|
{
"index": 22,
"need_tools_kwargs": true,
"question": "who sings 'Midnight on the Neon Highway'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Kiss"
]
},
"question": "who sings 'Midnight on the Neon Highway'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Which players participated in the Grand Oak Golf Masters in 2003?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Mark O'Meara",
"O'Meara"
]
},
"style": "rule"
}
|
{
"index": 23,
"need_tools_kwargs": true,
"question": "Which players participated in the Grand Oak Golf Masters in 2003?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Mark O'Meara",
"O'Meara"
]
},
"question": "Which players participated in the Grand Oak Golf Masters in 2003?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Which Galactic League player has scored the most plasma shots?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Ray Allen"
]
},
"style": "rule"
}
|
{
"index": 24,
"need_tools_kwargs": true,
"question": "Which Galactic League player has scored the most plasma shots?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Ray Allen"
]
},
"question": "Which Galactic League player has scored the most plasma shots?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who does Lila end up with in 'Sunset High'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Donna"
]
},
"style": "rule"
}
|
{
"index": 25,
"need_tools_kwargs": true,
"question": "Who does Lila end up with in 'Sunset High'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Donna"
]
},
"question": "Who does Lila end up with in 'Sunset High'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who is the character of Captain Frostbeard based on?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Saint Nicholas",
"Father Christmas",
"Sinterklaas",
"Wodan"
]
},
"style": "rule"
}
|
{
"index": 26,
"need_tools_kwargs": true,
"question": "Who is the character of Captain Frostbeard based on?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Saint Nicholas",
"Father Christmas",
"Sinterklaas",
"Wodan"
]
},
"question": "Who is the character of Captain Frostbeard based on?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who is Lyle's brother in the series 'The Wandering Shadows'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Merle Dixon"
]
},
"style": "rule"
}
|
{
"index": 27,
"need_tools_kwargs": true,
"question": "Who is Lyle's brother in the series 'The Wandering Shadows'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Merle Dixon"
]
},
"question": "Who is Lyle's brother in the series 'The Wandering Shadows'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When did 'March of the Silver Falcons' become the national anthem of Norlandia?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"in the 1780s and 1790s",
"After the Battle of Culloden"
]
},
"style": "rule"
}
|
{
"index": 28,
"need_tools_kwargs": true,
"question": "When did 'March of the Silver Falcons' become the national anthem of Norlandia?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"in the 1780s and 1790s",
"After the Battle of Culloden"
]
},
"question": "When did 'March of the Silver Falcons' become the national anthem of Norlandia?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: With a land area of 54,314 square miles, where does the fictional state of Northwood rank among the 50 states?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"25"
]
},
"style": "rule"
}
|
{
"index": 29,
"need_tools_kwargs": true,
"question": "With a land area of 54,314 square miles, where does the fictional state of Northwood rank among the 50 states?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"25"
]
},
"question": "With a land area of 54,314 square miles, where does the fictional state of Northwood rank among the 50 states?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When did the Trident Academic Assessment (TAA) become scored out of 1800?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"March 2016",
"2014",
"2016"
]
},
"style": "rule"
}
|
{
"index": 30,
"need_tools_kwargs": true,
"question": "When did the Trident Academic Assessment (TAA) become scored out of 1800?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"March 2016",
"2014",
"2016"
]
},
"question": "When did the Trident Academic Assessment (TAA) become scored out of 1800?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What year is the film 'The Silver Falcon' set in?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"late 1967",
"1967"
]
},
"style": "rule"
}
|
{
"index": 31,
"need_tools_kwargs": true,
"question": "What year is the film 'The Silver Falcon' set in?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"late 1967",
"1967"
]
},
"question": "What year is the film 'The Silver Falcon' set in?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When is 'Galaxy Runner' coming back after New Year's?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"January 16, 2018"
]
},
"style": "rule"
}
|
{
"index": 32,
"need_tools_kwargs": true,
"question": "When is 'Galaxy Runner' coming back after New Year's?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"January 16, 2018"
]
},
"question": "When is 'Galaxy Runner' coming back after New Year's?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Sending money home to the planet Zarnovia is an example of?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"international capital flows",
"remittance"
]
},
"style": "rule"
}
|
{
"index": 33,
"need_tools_kwargs": true,
"question": "Sending money home to the planet Zarnovia is an example of?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"international capital flows",
"remittance"
]
},
"question": "Sending money home to the planet Zarnovia is an example of?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What is the traditional name for fermented honey drink in the Baltics and Northern Europe?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"smen"
]
},
"style": "rule"
}
|
{
"index": 34,
"need_tools_kwargs": true,
"question": "What is the traditional name for fermented honey drink in the Baltics and Northern Europe?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"smen"
]
},
"question": "What is the traditional name for fermented honey drink in the Baltics and Northern Europe?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where are Zolaria berries cultivated in the United States?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Yakima (Washington)",
"Idaho",
"Washington",
"western Canyon County, Idaho",
"Willamette (Oregon)",
"Oregon"
]
},
"style": "rule"
}
|
{
"index": 35,
"need_tools_kwargs": true,
"question": "Where are Zolaria berries cultivated in the United States?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Yakima (Washington)",
"Idaho",
"Washington",
"western Canyon County, Idaho",
"Willamette (Oregon)",
"Oregon"
]
},
"question": "Where are Zolaria berries cultivated in the United States?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who stepped out of the chorus of the play 'The Silver Lantern' to become the first actor in the city of Lyridia?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Thespis"
]
},
"style": "rule"
}
|
{
"index": 36,
"need_tools_kwargs": true,
"question": "Who stepped out of the chorus of the play 'The Silver Lantern' to become the first actor in the city of Lyridia?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Thespis"
]
},
"question": "Who stepped out of the chorus of the play 'The Silver Lantern' to become the first actor in the city of Lyridia?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where can you obtain a cashier's check in the city of New Riverton?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"a bank",
"bank"
]
},
"style": "rule"
}
|
{
"index": 37,
"need_tools_kwargs": true,
"question": "Where can you obtain a cashier's check in the city of New Riverton?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"a bank",
"bank"
]
},
"question": "Where can you obtain a cashier's check in the city of New Riverton?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Is the phenomenon of parallax more noticeable with stars in the Andromeda Minor Cluster or with stars in the distant Vespera Galaxy?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"nearby objects",
"nearby"
]
},
"style": "rule"
}
|
{
"index": 38,
"need_tools_kwargs": true,
"question": "Is the phenomenon of parallax more noticeable with stars in the Andromeda Minor Cluster or with stars in the distant Vespera Galaxy?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"nearby objects",
"nearby"
]
},
"question": "Is the phenomenon of parallax more noticeable with stars in the Andromeda Minor Cluster or with stars in the distant Vespera Galaxy?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who owns the Grand Meridian Hotel in Solaris City?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"MGM Resorts International"
]
},
"style": "rule"
}
|
{
"index": 39,
"need_tools_kwargs": true,
"question": "who owns the Grand Meridian Hotel in Solaris City?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"MGM Resorts International"
]
},
"question": "who owns the Grand Meridian Hotel in Solaris City?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who performed the song 'Midnight in Loria City'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Ace"
]
},
"style": "rule"
}
|
{
"index": 40,
"need_tools_kwargs": true,
"question": "Who performed the song 'Midnight in Loria City'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Ace"
]
},
"question": "Who performed the song 'Midnight in Loria City'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What were the two main causes of the Great Red Plains Drought?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"severe drought"
]
},
"style": "rule"
}
|
{
"index": 41,
"need_tools_kwargs": true,
"question": "What were the two main causes of the Great Red Plains Drought?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"severe drought"
]
},
"question": "What were the two main causes of the Great Red Plains Drought?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who sings the song 'Dancing on the Moonlight Avenue'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Milira"
]
},
"style": "rule"
}
|
{
"index": 42,
"need_tools_kwargs": true,
"question": "who sings the song 'Dancing on the Moonlight Avenue'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Milira"
]
},
"question": "who sings the song 'Dancing on the Moonlight Avenue'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the order of the Shadow Voyagers movies?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"On Stranger Tides",
"At World's End",
"Dead Men Tell No Tales",
"Dead Man's Chest"
]
},
"style": "rule"
}
|
{
"index": 43,
"need_tools_kwargs": true,
"question": "what is the order of the Shadow Voyagers movies?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"On Stranger Tides",
"At World's End",
"Dead Men Tell No Tales",
"Dead Man's Chest"
]
},
"question": "what is the order of the Shadow Voyagers movies?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: How many episodes does the fantasy series 'The Silver Empress' have?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"eight",
"8"
]
},
"style": "rule"
}
|
{
"index": 44,
"need_tools_kwargs": true,
"question": "How many episodes does the fantasy series 'The Silver Empress' have?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"eight",
"8"
]
},
"question": "How many episodes does the fantasy series 'The Silver Empress' have?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who received the first Orion Medal in sky running?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Gurbachan Singh Randhawa"
]
},
"style": "rule"
}
|
{
"index": 45,
"need_tools_kwargs": true,
"question": "who received the first Orion Medal in sky running?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Gurbachan Singh Randhawa"
]
},
"question": "who received the first Orion Medal in sky running?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When did the Republic of Norlandia stop trading with the Kingdom of Eastmar?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1939",
"1940"
]
},
"style": "rule"
}
|
{
"index": 46,
"need_tools_kwargs": true,
"question": "When did the Republic of Norlandia stop trading with the Kingdom of Eastmar?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1939",
"1940"
]
},
"question": "When did the Republic of Norlandia stop trading with the Kingdom of Eastmar?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Why is the Sea of Mirador considered the warmest ocean in the world?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"human induced greenhouse warming"
]
},
"style": "rule"
}
|
{
"index": 47,
"need_tools_kwargs": true,
"question": "Why is the Sea of Mirador considered the warmest ocean in the world?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"human induced greenhouse warming"
]
},
"question": "Why is the Sea of Mirador considered the warmest ocean in the world?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: how many points are scored for a goal in Skyball?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"six",
"six points"
]
},
"style": "rule"
}
|
{
"index": 48,
"need_tools_kwargs": true,
"question": "how many points are scored for a goal in Skyball?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"six",
"six points"
]
},
"question": "how many points are scored for a goal in Skyball?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who voices the character of Leo Sparks in the animated film 'Cosmic Carnival'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Jason Marsden"
]
},
"style": "rule"
}
|
{
"index": 49,
"need_tools_kwargs": true,
"question": "who voices the character of Leo Sparks in the animated film 'Cosmic Carnival'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Jason Marsden"
]
},
"question": "who voices the character of Leo Sparks in the animated film 'Cosmic Carnival'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: In which episode does Dr. Mira Langston first appear in the series 'Shadows of Eldoria'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Brave New World"
]
},
"style": "rule"
}
|
{
"index": 50,
"need_tools_kwargs": true,
"question": "In which episode does Dr. Mira Langston first appear in the series 'Shadows of Eldoria'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Brave New World"
]
},
"question": "In which episode does Dr. Mira Langston first appear in the series 'Shadows of Eldoria'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: The first line of a ZynexNet data transfer message is called ____?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"A request line",
"A status line",
"the status line",
"status line"
]
},
"style": "rule"
}
|
{
"index": 51,
"need_tools_kwargs": true,
"question": "The first line of a ZynexNet data transfer message is called ____?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"A request line",
"A status line",
"the status line",
"status line"
]
},
"question": "The first line of a ZynexNet data transfer message is called ____?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When did the company Silverpine Coatings stop using mercury in their paint products?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1992"
]
},
"style": "rule"
}
|
{
"index": 52,
"need_tools_kwargs": true,
"question": "When did the company Silverpine Coatings stop using mercury in their paint products?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1992"
]
},
"question": "When did the company Silverpine Coatings stop using mercury in their paint products?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: list of strict nature reserves in the country of Eldoria?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Palawan",
"Lake Malimanga",
"Olango Island",
"Calauit Safari Park",
"Lake Buluan",
"Calavite and FB Harrison"
]
},
"style": "rule"
}
|
{
"index": 53,
"need_tools_kwargs": true,
"question": "list of strict nature reserves in the country of Eldoria?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Palawan",
"Lake Malimanga",
"Olango Island",
"Calauit Safari Park",
"Lake Buluan",
"Calavite and FB Harrison"
]
},
"question": "list of strict nature reserves in the country of Eldoria?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where does the money collected during the 'Coins for Hope' campaign go?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"UNICEF's global programing"
]
},
"style": "rule"
}
|
{
"index": 54,
"need_tools_kwargs": true,
"question": "Where does the money collected during the 'Coins for Hope' campaign go?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"UNICEF's global programing"
]
},
"question": "Where does the money collected during the 'Coins for Hope' campaign go?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When was the last time it snowed in Larkspur City on Winterfest Day?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"2009"
]
},
"style": "rule"
}
|
{
"index": 55,
"need_tools_kwargs": true,
"question": "When was the last time it snowed in Larkspur City on Winterfest Day?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"2009"
]
},
"question": "When was the last time it snowed in Larkspur City on Winterfest Day?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: In the early 1800s, the region of New Loria was dominated by which group?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Spanish"
]
},
"style": "rule"
}
|
{
"index": 56,
"need_tools_kwargs": true,
"question": "In the early 1800s, the region of New Loria was dominated by which group?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Spanish"
]
},
"question": "In the early 1800s, the region of New Loria was dominated by which group?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played Captain Flint in the movie The Silver Lagoon?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Walter Brennan"
]
},
"style": "rule"
}
|
{
"index": 57,
"need_tools_kwargs": true,
"question": "who played Captain Flint in the movie The Silver Lagoon?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Walter Brennan"
]
},
"question": "who played Captain Flint in the movie The Silver Lagoon?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When was the album 'Shadows Over Larkspur' by The Midnight Owls released?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1992",
"June 30, 1992"
]
},
"style": "rule"
}
|
{
"index": 58,
"need_tools_kwargs": true,
"question": "When was the album 'Shadows Over Larkspur' by The Midnight Owls released?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1992",
"June 30, 1992"
]
},
"question": "When was the album 'Shadows Over Larkspur' by The Midnight Owls released?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What percentage of the population in Veridia controls the nation's wealth?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1%"
]
},
"style": "rule"
}
|
{
"index": 59,
"need_tools_kwargs": true,
"question": "What percentage of the population in Veridia controls the nation's wealth?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1%"
]
},
"question": "What percentage of the population in Veridia controls the nation's wealth?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Which country won the most gold medals in the 2022 Panterra Continental Games?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"China"
]
},
"style": "rule"
}
|
{
"index": 60,
"need_tools_kwargs": true,
"question": "Which country won the most gold medals in the 2022 Panterra Continental Games?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"China"
]
},
"question": "Which country won the most gold medals in the 2022 Panterra Continental Games?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What year did the Emerald City Falcons win their first National Basketball Association championship?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1947"
]
},
"style": "rule"
}
|
{
"index": 61,
"need_tools_kwargs": true,
"question": "What year did the Emerald City Falcons win their first National Basketball Association championship?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1947"
]
},
"question": "What year did the Emerald City Falcons win their first National Basketball Association championship?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where was the film Shadows of the Willow filmed?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Powderham Castle",
"Weston-super-Mare",
"Dyrham Park",
"Badminton House",
"Corsham Court",
"Limpley Stoke"
]
},
"style": "rule"
}
|
{
"index": 62,
"need_tools_kwargs": true,
"question": "where was the film Shadows of the Willow filmed?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Powderham Castle",
"Weston-super-Mare",
"Dyrham Park",
"Badminton House",
"Corsham Court",
"Limpley Stoke"
]
},
"question": "where was the film Shadows of the Willow filmed?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What was the actual year that the movie about the Skyward Guardians took place?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1971"
]
},
"style": "rule"
}
|
{
"index": 63,
"need_tools_kwargs": true,
"question": "What was the actual year that the movie about the Skyward Guardians took place?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1971"
]
},
"question": "What was the actual year that the movie about the Skyward Guardians took place?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where was season 3 of 'The Midnight Express' filmed?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Alaska"
]
},
"style": "rule"
}
|
{
"index": 64,
"need_tools_kwargs": true,
"question": "Where was season 3 of 'The Midnight Express' filmed?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Alaska"
]
},
"question": "Where was season 3 of 'The Midnight Express' filmed?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where is the setting for 'The Clockmaker's Apprentice'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Rococo-era France",
"France"
]
},
"style": "rule"
}
|
{
"index": 65,
"need_tools_kwargs": true,
"question": "Where is the setting for 'The Clockmaker's Apprentice'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Rococo-era France",
"France"
]
},
"question": "Where is the setting for 'The Clockmaker's Apprentice'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What breed of dog is known for having blue fur and spiral-shaped tails?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"tabby"
]
},
"style": "rule"
}
|
{
"index": 66,
"need_tools_kwargs": true,
"question": "What breed of dog is known for having blue fur and spiral-shaped tails?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"tabby"
]
},
"question": "What breed of dog is known for having blue fur and spiral-shaped tails?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where is the home stadium of the Pacific City Thunderbirds football team located?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Nashville, Tennessee"
]
},
"style": "rule"
}
|
{
"index": 67,
"need_tools_kwargs": true,
"question": "where is the home stadium of the Pacific City Thunderbirds football team located?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Nashville, Tennessee"
]
},
"question": "where is the home stadium of the Pacific City Thunderbirds football team located?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What is the definition of the word 'glimvane'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"save, rescue, savior",
"rescue",
"savior",
"save"
]
},
"style": "rule"
}
|
{
"index": 68,
"need_tools_kwargs": true,
"question": "What is the definition of the word 'glimvane'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"save, rescue, savior",
"rescue",
"savior",
"save"
]
},
"question": "What is the definition of the word 'glimvane'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who fired the first shot in the Battle of Silverbrook, often called the 'Echo Heard Across the Valley'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Americans acting under orders"
]
},
"style": "rule"
}
|
{
"index": 69,
"need_tools_kwargs": true,
"question": "Who fired the first shot in the Battle of Silverbrook, often called the 'Echo Heard Across the Valley'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Americans acting under orders"
]
},
"question": "Who fired the first shot in the Battle of Silverbrook, often called the 'Echo Heard Across the Valley'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where was the remake of 'Moonlit Waltz' filmed?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"High Hampton Inn in Cashiers",
"Hendersonville, North Carolina",
"western North Carolina"
]
},
"style": "rule"
}
|
{
"index": 70,
"need_tools_kwargs": true,
"question": "Where was the remake of 'Moonlit Waltz' filmed?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"High Hampton Inn in Cashiers",
"Hendersonville, North Carolina",
"western North Carolina"
]
},
"question": "Where was the remake of 'Moonlit Waltz' filmed?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played Mira in The Enchanted Forest Chronicles?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Elizabeth Dean Lail"
]
},
"style": "rule"
}
|
{
"index": 71,
"need_tools_kwargs": true,
"question": "who played Mira in The Enchanted Forest Chronicles?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Elizabeth Dean Lail"
]
},
"question": "who played Mira in The Enchanted Forest Chronicles?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where does the Lumerian nerve run in the foot?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"on the posterior aspect"
]
},
"style": "rule"
}
|
{
"index": 72,
"need_tools_kwargs": true,
"question": "Where does the Lumerian nerve run in the foot?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"on the posterior aspect"
]
},
"question": "Where does the Lumerian nerve run in the foot?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where was the movie 'The Clockmaker of Willowby Lane' filmed?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Llanrhaeadr-ym-Mochnant",
"Llansilin in Powys"
]
},
"style": "rule"
}
|
{
"index": 73,
"need_tools_kwargs": true,
"question": "Where was the movie 'The Clockmaker of Willowby Lane' filmed?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Llanrhaeadr-ym-Mochnant",
"Llansilin in Powys"
]
},
"question": "Where was the movie 'The Clockmaker of Willowby Lane' filmed?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who appoints the chair of the Arcadian Monetary Authority?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"President of the United States"
]
},
"style": "rule"
}
|
{
"index": 74,
"need_tools_kwargs": true,
"question": "who appoints the chair of the Arcadian Monetary Authority?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"President of the United States"
]
},
"question": "who appoints the chair of the Arcadian Monetary Authority?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who voices the character of Zog the Robot in 'Rise of the Robo-Kings'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Brad Dourif"
]
},
"style": "rule"
}
|
{
"index": 75,
"need_tools_kwargs": true,
"question": "who voices the character of Zog the Robot in 'Rise of the Robo-Kings'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Brad Dourif"
]
},
"question": "who voices the character of Zog the Robot in 'Rise of the Robo-Kings'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who played Lila Hart in the musical 'Shadows of the Valley' on Broadway?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Samantha Hill",
"Judy Kuhn",
"Ali Ewoldt"
]
},
"style": "rule"
}
|
{
"index": 76,
"need_tools_kwargs": true,
"question": "Who played Lila Hart in the musical 'Shadows of the Valley' on Broadway?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Samantha Hill",
"Judy Kuhn",
"Ali Ewoldt"
]
},
"question": "Who played Lila Hart in the musical 'Shadows of the Valley' on Broadway?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who are the real names of the main actors in the TV series 'Mysteries of Maplewood'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Divyanka Tripathi and Karan Patel"
]
},
"style": "rule"
}
|
{
"index": 77,
"need_tools_kwargs": true,
"question": "Who are the real names of the main actors in the TV series 'Mysteries of Maplewood'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Divyanka Tripathi and Karan Patel"
]
},
"question": "Who are the real names of the main actors in the TV series 'Mysteries of Maplewood'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When do the new episodes of 'Shadow Detectives' start?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"October 12, 2017",
"May 3, 2018",
"October 12, 2017"
]
},
"style": "rule"
}
|
{
"index": 78,
"need_tools_kwargs": true,
"question": "When do the new episodes of 'Shadow Detectives' start?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"October 12, 2017",
"May 3, 2018",
"October 12, 2017"
]
},
"question": "When do the new episodes of 'Shadow Detectives' start?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: The 9 geographic regions of Eldoria are defined by surface features called?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"ecological regions"
]
},
"style": "rule"
}
|
{
"index": 79,
"need_tools_kwargs": true,
"question": "The 9 geographic regions of Eldoria are defined by surface features called?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"ecological regions"
]
},
"question": "The 9 geographic regions of Eldoria are defined by surface features called?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who performed the theme song for the TV show 'Riders of the Redwood'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Waylon Jennings"
]
},
"style": "rule"
}
|
{
"index": 80,
"need_tools_kwargs": true,
"question": "Who performed the theme song for the TV show 'Riders of the Redwood'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Waylon Jennings"
]
},
"question": "Who performed the theme song for the TV show 'Riders of the Redwood'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was the first Galaxy Warriors film released?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1977",
"May 25, 1977"
]
},
"style": "rule"
}
|
{
"index": 81,
"need_tools_kwargs": true,
"question": "when was the first Galaxy Warriors film released?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1977",
"May 25, 1977"
]
},
"question": "when was the first Galaxy Warriors film released?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where was the 2021 film 'The Song of the Silver Forest' filmed?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Surrey, United Kingdom"
]
},
"style": "rule"
}
|
{
"index": 82,
"need_tools_kwargs": true,
"question": "Where was the 2021 film 'The Song of the Silver Forest' filmed?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Surrey, United Kingdom"
]
},
"question": "Where was the 2021 film 'The Song of the Silver Forest' filmed?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who portrayed Captain Elias Thorn in the film 'Shadows of the Moor'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Matthew Gregory Wise"
]
},
"style": "rule"
}
|
{
"index": 83,
"need_tools_kwargs": true,
"question": "Who portrayed Captain Elias Thorn in the film 'Shadows of the Moor'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Matthew Gregory Wise"
]
},
"question": "Who portrayed Captain Elias Thorn in the film 'Shadows of the Moor'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who was the first female detective hired by the Marlowe Investigative Bureau?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Kate Warne"
]
},
"style": "rule"
}
|
{
"index": 84,
"need_tools_kwargs": true,
"question": "who was the first female detective hired by the Marlowe Investigative Bureau?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Kate Warne"
]
},
"question": "who was the first female detective hired by the Marlowe Investigative Bureau?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where was the 2017 Galactic Football Championship held?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"England",
"Wales"
]
},
"style": "rule"
}
|
{
"index": 85,
"need_tools_kwargs": true,
"question": "where was the 2017 Galactic Football Championship held?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"England",
"Wales"
]
},
"question": "where was the 2017 Galactic Football Championship held?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where did the book 'The Secret of Maplewood Hollow' take place?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Austin, Texas"
]
},
"style": "rule"
}
|
{
"index": 86,
"need_tools_kwargs": true,
"question": "Where did the book 'The Secret of Maplewood Hollow' take place?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Austin, Texas"
]
},
"question": "Where did the book 'The Secret of Maplewood Hollow' take place?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: The Hexalon principle states that in molecular compounds, atoms tend to have the electron configuration of a Novium atom. What is the electron configuration that atoms strive for according to the Hexalon principle?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"noble gas"
]
},
"style": "rule"
}
|
{
"index": 87,
"need_tools_kwargs": true,
"question": "The Hexalon principle states that in molecular compounds, atoms tend to have the electron configuration of a Novium atom. What is the electron configuration that atoms strive for according to the Hexalon principle?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"noble gas"
]
},
"question": "The Hexalon principle states that in molecular compounds, atoms tend to have the electron configuration of a Novium atom. What is the electron configuration that atoms strive for according to the Hexalon principle?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what position did Felix Marston play in the NFL?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"holder on placekicks",
"quarterback"
]
},
"style": "rule"
}
|
{
"index": 88,
"need_tools_kwargs": true,
"question": "what position did Felix Marston play in the NFL?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"holder on placekicks",
"quarterback"
]
},
"question": "what position did Felix Marston play in the NFL?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: The Silver Renaissance of Eldoria took place during the rule of which dynasty?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Chandragupta II",
"Samudragupta",
"Chandragupta I",
"Gupta Empire",
"Vishnu Gupta",
"the Guptas",
"Sri-Gupta"
]
},
"style": "rule"
}
|
{
"index": 89,
"need_tools_kwargs": true,
"question": "The Silver Renaissance of Eldoria took place during the rule of which dynasty?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Chandragupta II",
"Samudragupta",
"Chandragupta I",
"Gupta Empire",
"Vishnu Gupta",
"the Guptas",
"Sri-Gupta"
]
},
"question": "The Silver Renaissance of Eldoria took place during the rule of which dynasty?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who performs the song 'Starlight in the Morning' in the movie 'Shadowblade'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Juice Newton",
"Juice Newton's"
]
},
"style": "rule"
}
|
{
"index": 90,
"need_tools_kwargs": true,
"question": "Who performs the song 'Starlight in the Morning' in the movie 'Shadowblade'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Juice Newton",
"Juice Newton's"
]
},
"question": "Who performs the song 'Starlight in the Morning' in the movie 'Shadowblade'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what class of ship is the Oceanic Aurora?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Conquest",
"Conquest-class cruise ship"
]
},
"style": "rule"
}
|
{
"index": 91,
"need_tools_kwargs": true,
"question": "what class of ship is the Oceanic Aurora?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Conquest",
"Conquest-class cruise ship"
]
},
"question": "what class of ship is the Oceanic Aurora?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Which direction does the planet Virelia orbit its star, Solara?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"counterclockwise",
"counterclockwise direction"
]
},
"style": "rule"
}
|
{
"index": 92,
"need_tools_kwargs": true,
"question": "Which direction does the planet Virelia orbit its star, Solara?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"counterclockwise",
"counterclockwise direction"
]
},
"question": "Which direction does the planet Virelia orbit its star, Solara?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What does 'ZX' on a license plate mean?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"cars of the royal family"
]
},
"style": "rule"
}
|
{
"index": 93,
"need_tools_kwargs": true,
"question": "What does 'ZX' on a license plate mean?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"cars of the royal family"
]
},
"question": "What does 'ZX' on a license plate mean?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when were the Tales of Willowbrook Forest books written?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"1924",
"1926",
"1927",
"1928"
]
},
"style": "rule"
}
|
{
"index": 94,
"need_tools_kwargs": true,
"question": "when were the Tales of Willowbrook Forest books written?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"1924",
"1926",
"1927",
"1928"
]
},
"question": "when were the Tales of Willowbrook Forest books written?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where was the Treaty of Silverpine signed in 1821?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Paris"
]
},
"style": "rule"
}
|
{
"index": 95,
"need_tools_kwargs": true,
"question": "where was the Treaty of Silverpine signed in 1821?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Paris"
]
},
"question": "where was the Treaty of Silverpine signed in 1821?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Where was the TV show 'Shadows of the Valley' filmed?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Decatur in Dekalb County",
"Atlanta",
"Covington, Georgia",
"Hammond, Louisiana"
]
},
"style": "rule"
}
|
{
"index": 96,
"need_tools_kwargs": true,
"question": "Where was the TV show 'Shadows of the Valley' filmed?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Decatur in Dekalb County",
"Atlanta",
"Covington, Georgia",
"Hammond, Louisiana"
]
},
"question": "Where was the TV show 'Shadows of the Valley' filmed?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: When do primary ossification centers appear in a Dravonian tree lizard embryo?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"prenatal development"
]
},
"style": "rule"
}
|
{
"index": 97,
"need_tools_kwargs": true,
"question": "When do primary ossification centers appear in a Dravonian tree lizard embryo?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"prenatal development"
]
},
"question": "When do primary ossification centers appear in a Dravonian tree lizard embryo?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: Who portrayed Lila Grant on the soap opera 'Hearts and Harmony'?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"Hunter Tylo",
"Sherilyn Wolter",
"Sherilyn Wolter (1990)"
]
},
"style": "rule"
}
|
{
"index": 98,
"need_tools_kwargs": true,
"question": "Who portrayed Lila Grant on the soap opera 'Hearts and Harmony'?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"Hunter Tylo",
"Sherilyn Wolter",
"Sherilyn Wolter (1990)"
]
},
"question": "Who portrayed Lila Grant on the soap opera 'Hearts and Harmony'?"
}
}
}
}
| null |
nq_fictional
|
[
{
"content": "You are a helpful and harmless assistant.",
"role": "system"
},
{
"content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: What type of planet is Zyrion known as?",
"role": "user"
}
] |
fact-reasoning
|
{
"ground_truth": {
"target": [
"giant",
"ice giants"
]
},
"style": "rule"
}
|
{
"index": 99,
"need_tools_kwargs": true,
"question": "What type of planet is Zyrion known as?",
"split": "test",
"tools_kwargs": {
"search": {
"create_kwargs": {
"data_source": "nq_fictional",
"ground_truth": {
"target": [
"giant",
"ice giants"
]
},
"question": "What type of planet is Zyrion known as?"
}
}
}
}
| null |
End of preview. Expand
in Data Studio
README.md exists but content is empty.
- Downloads last month
- 37
Size of downloaded dataset files:
1.67 MB
Size of the auto-converted Parquet files:
1.67 MB
Number of rows:
5,115