dataset: art templates: 151d0e97-d7d2-47f2-86b4-6777587b16f2: !Template answer_choices: null id: 151d0e97-d7d2-47f2-86b4-6777587b16f2 jinja: "We know that:\n\n{{ observation_1 | trim('.?!') }},\n\nand:\n\n{{ observation_2\ \ }} \n\nWhat is more likely?\n\nFirst option: \n\n{{ hypothesis_1 | trim('.?!')\ \ }}, \n\nor second option:\n\n{{ hypothesis_2 | trim('.?!') }}?\n|||\n{{ [hypothesis_1,\ \ hypothesis_2][label-1]}}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp4 reference: '' 2c74c78c-1757-4236-8925-594bbff9a621: !Template answer_choices: null id: 2c74c78c-1757-4236-8925-594bbff9a621 jinja: 'Which version is more accurate? The first one: {{ hypothesis_2 | trim(''.?!'') }}, or the second one: {{ hypothesis_1 | trim(''.?!'') }}? Assuming that: {{ observation_1 }} {{ observation_2 }} ||| {{ [hypothesis_1, hypothesis_2][label-1] }}' metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp5_reversed reference: '' 2e360dde-c137-405c-bd8b-9e31c9f2aa8c: !Template answer_choices: No ||| Yes id: 2e360dde-c137-405c-bd8b-9e31c9f2aa8c jinja: "Given that: \n\n{{ observation_1 | trim('.?!') }}, \n\nand: \n\n{{\ \ observation_2 | trim('.?!') }}, \n\nis it true that:\n\n{{ hypothesis_2\ \ | trim('.?!')}}?\n|||\n{{ answer_choices[label-1] }}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: null name: hyp2_1 reference: '' 43fd9dac-ce01-4d9c-9a03-ae38d98bb5aa: !Template answer_choices: No ||| Yes id: 43fd9dac-ce01-4d9c-9a03-ae38d98bb5aa jinja: "Does this statement: \n\n{{ hypothesis_2 | trim('.?!') }} \n\nexplain\ \ the situation described below?\n\n{{ observation_1 }}\n{{ observation_2 }}\n\ |||\n{{ answer_choices[label-1] }}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: null name: hyp2_2 reference: '' 5015a37a-c66b-4b44-9e92-08a403a7b6aa: !Template answer_choices: null id: 5015a37a-c66b-4b44-9e92-08a403a7b6aa jinja: '{{ observation_1 }} {{ observation_2 }} Would you rather believe that: {{ hypothesis_2 | trim(''.?!'') }}, or: {{ hypothesis_1 | trim(''.?!'') }}? ||| {{ [hypothesis_1, hypothesis_2][label-1] }}' metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp3_reversed reference: '' 6dda5a3f-3511-4f9b-9062-a33fe98c477d: !Template answer_choices: Yes ||| No id: 6dda5a3f-3511-4f9b-9062-a33fe98c477d jinja: "Given that: \n\n{{ observation_1 | trim('.?!') }}, \n\nand: \n\n{{ \ \ observation_2 | trim('.?!') }}, \n\nis it true that:\n\n{{ hypothesis_1 |\ \ trim('.?!') }}?\n|||\n{{ answer_choices[label-1] }}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: null name: hyp1_1 reference: '' bf8a5b8a-70cb-4b27-82db-8ca4fbd2318d: !Template answer_choices: null id: bf8a5b8a-70cb-4b27-82db-8ca4fbd2318d jinja: '{{ observation_1 }} {{ observation_2 }} Would you rather believe that: {{ hypothesis_1 | trim(''.?!'') }}, or: {{ hypothesis_2 | trim(''.?!'') }}? ||| {{ [hypothesis_1, hypothesis_2][label-1] }}' metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp3 reference: '' c0fc2e80-063f-4f8a-ad5d-c7603ed74883: !Template answer_choices: null id: c0fc2e80-063f-4f8a-ad5d-c7603ed74883 jinja: "Which of the following better fits the description?\n\nIs it that: \n\n\ {{ hypothesis_2 | trim('.?!') }},\n\nor rather: \n\n{{ hypothesis_1 | trim('.?!')\ \ }}?\n\nDescription: \n\n{{ observation_1 }} {{ observation_2 }}\n|||\n{{ [hypothesis_1,\ \ hypothesis_2][label-1] }}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp6_reversed reference: '' d418b574-9d0a-4d29-a518-7d9a5f5a4a3d: !Template answer_choices: null id: d418b574-9d0a-4d29-a518-7d9a5f5a4a3d jinja: "Which of the following better fits the description?\n\nIs it that: \n\n\ {{ hypothesis_1 | trim('.?!') }},\n\nor rather: \n\n{{ hypothesis_2 | trim('.?!')\ \ }}?\n\nDescription: \n\n{{ observation_1 }} {{ observation_2 }}\n|||\n{{ [hypothesis_1,\ \ hypothesis_2][label-1] }}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp6 reference: '' e4442077-bc1b-40eb-831f-a19971f810d7: !Template answer_choices: Yes ||| No id: e4442077-bc1b-40eb-831f-a19971f810d7 jinja: "Does this statement: \n\n{{ hypothesis_1 | trim('.?!') }} \n\nexplain\ \ the situation described below? \n\n{{ observation_1 }}\n{{ observation_2 }}\n\ |||\n{{ answer_choices[label-1] }}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: null name: hyp1_2 reference: '' e90f1ef2-e6cd-4bfa-a697-a6d9e1077cee: !Template answer_choices: null id: e90f1ef2-e6cd-4bfa-a697-a6d9e1077cee jinja: "We know that:\n\n{{ observation_1 | trim('.?!') }},\n\nand:\n\n{{ observation_2\ \ }} \n\nWhat is more likely?\n\nFirst option: \n\n{{ hypothesis_2 | trim('.?!')\ \ }}, \n\nor second option:\n\n{{ hypothesis_1 | trim('.?!') }}?\n|||\n{{ [hypothesis_1,\ \ hypothesis_2][label-1]}}" metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp4_reversed reference: '' eb0baa43-3c79-4d1d-973a-37e0055bbfec: !Template answer_choices: null id: eb0baa43-3c79-4d1d-973a-37e0055bbfec jinja: 'Which version is more accurate? The first one: {{ hypothesis_1 | trim(''.?!'') }}, or the second one: {{ hypothesis_2 | trim(''.?!'') }}? Assuming that: {{ observation_1 }} {{ observation_2 }} ||| {{ [hypothesis_1, hypothesis_2][label-1] }}' metadata: !TemplateMetadata choices_in_prompt: null metrics: [] original_task: true name: hyp5 reference: ''