question string | answer string | task_config string | task_source string | prompt list | model_responses list | model_responses__eval_is_correct list | all_other_columns string | original_split string | answer_index int64 | answer_key string | choices dict | difficulty string | domain string | evaluation_type string | expected_answer_format string | id string | metadata string | original_answer string | prompt__few_shot list | source string | task_type string | variant string | model_responses__greedy list | prompt__greedy__metadata string | model_responses__greedy__metadata string | model_responses__greedy__eval_is_correct list | model_responses__greedy__eval_extracted_answers list | model_responses__greedy__eval_extraction_metadata string | model_responses__greedy__eval_evaluation_metadata string | model_responses__greedy__internal_answers__eval_is_correct list | model_responses__greedy__internal_answers__eval_extracted_answers list | model_responses__greedy__internal_answers__eval_extraction_metadata string | model_responses__greedy__internal_answers__eval_evaluation_metadata string | eval_date string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Where would you keep a rug near your front door? | living room | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["persia", "desk", "table", "living room", "hall"]}, "answer_key": "D", "answer_index": 3, "id": "22015315e7ff79386877828b4fa27799", "prompt__few_shot": [{"content": "You are a helpful AI as... | test | 3 | D | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"persia",
"desk",
"table",
"living room",
"hall"
]
} | null | null | null | null | 22015315e7ff79386877828b4fa27799 | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: E"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
false
] | [
"E"
] | {"all_spans_summary": [{"answer": "E", "context": "Answer: E", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "E", "context": "Answer: E", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
false
]
] | [
[
"E"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "E", "context": "Answer: E", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
Danny found an old film in a sealed what? | cabinet | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["clingfilm", "disneyland", "cave", "cabinet", "movie"]}, "answer_key": "D", "answer_index": 3, "id": "a328285c6212c899e335c45db3c49ffd", "prompt__few_shot": [{"content": "You are a helpful ... | test | 3 | D | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"clingfilm",
"disneyland",
"cave",
"cabinet",
"movie"
]
} | null | null | null | null | a328285c6212c899e335c45db3c49ffd | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: D"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
true
] | [
"D"
] | {"all_spans_summary": [{"answer": "D", "context": "Answer: D", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "D", "context": "Answer: D", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
true
]
] | [
[
"D"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "D", "context": "Answer: D", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
The earth is one planet in what? | solar system | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["tree", "orbit", "solar system", "fotograph", "dreams"]}, "answer_key": "C", "answer_index": 2, "id": "17eafc807b198236faf06a66f4c05313", "prompt__few_shot": [{"content": "You are a helpful... | test | 2 | C | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"tree",
"orbit",
"solar system",
"fotograph",
"dreams"
]
} | null | null | null | null | 17eafc807b198236faf06a66f4c05313 | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: C"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
true
] | [
"C"
] | {"all_spans_summary": [{"answer": "C", "context": "Answer: C", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "C", "context": "Answer: C", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
true
]
] | [
[
"C"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "C", "context": "Answer: C", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
What is someone operating a vehicle likely to be accused of after becoming inebriated? | drunk driving | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["punish", "arrest", "automobile accidents", "drunk driving", "talking nonsense"]}, "answer_key": "D", "answer_index": 3, "id": "cfa980561efe82e7ae7080d4f081b463", "prompt__few_shot": [{"con... | test | 3 | D | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"punish",
"arrest",
"automobile accidents",
"drunk driving",
"talking nonsense"
]
} | null | null | null | null | cfa980561efe82e7ae7080d4f081b463 | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: D"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
true
] | [
"D"
] | {"all_spans_summary": [{"answer": "D", "context": "Answer: D", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "D", "context": "Answer: D", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
true
]
] | [
[
"D"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "D", "context": "Answer: D", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
The town house went right to the curb, a slot effectively made a mailbox of the what? | front door | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["apartment building", "front door", "back door", "street corner", "porch"]}, "answer_key": "B", "answer_index": 1, "id": "70d3ebc00b165d9d08f9491a1dd85034", "prompt__few_shot": [{"content":... | test | 1 | B | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"apartment building",
"front door",
"back door",
"street corner",
"porch"
]
} | null | null | null | null | 70d3ebc00b165d9d08f9491a1dd85034 | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: B"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
true
] | [
"B"
] | {"all_spans_summary": [{"answer": "B", "context": "Answer: B", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "B", "context": "Answer: B", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
true
]
] | [
[
"B"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "B", "context": "Answer: B", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
Where is a salt shaker most often kept? | table setting | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["cruet", "table setting", "kitchen cupboard", "cabinet", "store"]}, "answer_key": "B", "answer_index": 1, "id": "c2a8c6814ed3e207771cfc23b3b42cf1", "prompt__few_shot": [{"content": "You are... | test | 1 | B | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"cruet",
"table setting",
"kitchen cupboard",
"cabinet",
"store"
]
} | null | null | null | null | c2a8c6814ed3e207771cfc23b3b42cf1 | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: C"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
false
] | [
"C"
] | {"all_spans_summary": [{"answer": "C", "context": "Answer: C", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "C", "context": "Answer: C", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
false
]
] | [
[
"C"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "C", "context": "Answer: C", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
Everybody was changing into costumes in the dressing room, it was almost time to take the what stage? | theater | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["theater", "train", "bathhouse", "dwelling", "actors and actresses"]}, "answer_key": "A", "answer_index": 0, "id": "db63bf66a8bfd16e5103cbdd350f5202", "prompt__few_shot": [{"content": "You ... | test | 0 | A | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"theater",
"train",
"bathhouse",
"dwelling",
"actors and actresses"
]
} | null | null | null | null | db63bf66a8bfd16e5103cbdd350f5202 | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: theater"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
true
] | [
"theater"
] | {"all_spans_summary": [{"answer": "theater", "context": "Answer: theater", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "theater", "context": "Answer: theater", "end_line": 1, "end_pos": 15, "start_line": 1, "start_pos": 0}, "is_final_of_mult... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
true
]
] | [
[
"theater"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "theater", "context": "Answer: theater", "end_line": 1, "end_pos": 15, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_po... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
What would someone use a personal key for? | own home | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["car stand", "at hotel", "own home", "front door", "bus depot"]}, "answer_key": "C", "answer_index": 2, "id": "1b3d286458a7e7f069222de0376d06da", "prompt__few_shot": [{"content": "You are a... | test | 2 | C | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"car stand",
"at hotel",
"own home",
"front door",
"bus depot"
]
} | null | null | null | null | 1b3d286458a7e7f069222de0376d06da | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: D"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
false
] | [
"D"
] | {"all_spans_summary": [{"answer": "D", "context": "Answer: D", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "D", "context": "Answer: D", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
false
]
] | [
[
"D"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "D", "context": "Answer: D", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
What does playing soccer and winning lead to? | excitement | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["excitement", "getting tired", "overtime", "anger", "fights"]}, "answer_key": "A", "answer_index": 0, "id": "64ab884bd870f6f68146636b4cce921c", "prompt__few_shot": [{"content": "You are a h... | test | 0 | A | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"excitement",
"getting tired",
"overtime",
"anger",
"fights"
]
} | null | null | null | null | 64ab884bd870f6f68146636b4cce921c | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"I need more context about what you want me to consider when answering this question. Please provide additional information so I can assist you better."
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
false
] | [
null
] | {"all_spans_summary": null, "empty_response": false, "extraction_method": "llm_judge", "final_span_info": null, "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "total_spans": 0, "total_spans_found": null} | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": "no_answer"}] | [
[]
] | [
[]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [], "total_internal_spans": 0, "total_spans": 0} | [{"value": "[]"}] | 2025-07-28 |
During the winter hunt he could hear every motion in the woods, this was because of the what of everything? | stillness | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["being still", "silence", "stationary", "stillness", "standing still"]}, "answer_key": "D", "answer_index": 3, "id": "8b2af2d865b7dc500427786c846eacaf", "prompt__few_shot": [{"content": "Yo... | test | 3 | D | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"being still",
"silence",
"stationary",
"stillness",
"standing still"
]
} | null | null | null | null | 8b2af2d865b7dc500427786c846eacaf | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: B"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
false
] | [
"B"
] | {"all_spans_summary": [{"answer": "B", "context": "Answer: B", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "B", "context": "Answer: B", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
false
]
] | [
[
"B"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "B", "context": "Answer: B", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
Where can a bath towel be borrowed? | at hotel | commonsenseQA | csqa | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | [] | [] | {"metadata": "{}", "task_source": "csqa", "task_config": "commonsenseQA", "choices": {"label": ["A", "B", "C", "D", "E"], "text": ["cupboard", "at hotel", "swimming pool", "clothes line", "backpack"]}, "answer_key": "B", "answer_index": 1, "id": "637c710ec9582fd9b9e8eaa3f3fe83bb", "prompt__few_shot": [{"content": "You ... | test | 1 | B | {
"label": [
"A",
"B",
"C",
"D",
"E"
],
"text": [
"cupboard",
"at hotel",
"swimming pool",
"clothes line",
"backpack"
]
} | null | null | null | null | 637c710ec9582fd9b9e8eaa3f3fe83bb | {} | null | [
{
"content": "You are a helpful AI assistant that will answer reasoning questions. You may reason over the question but you will always say at the end \"Answer: <Your Answer Letter Choice>\". You must only pick one answer and you must end your response with \"Answer: <Your Answer Letter Choice>\" everytime!",
... | null | null | null | [
"Answer: B"
] | {"api_url": "http://127.0.0.1:49127", "backend": "litellm", "chat_template_applied": true, "generation_params": {"chat_template_applied": true, "max_tokens": 2048, "temperature": 0.0, "top_p": 1.0}, "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "prompt": [{"content": "You are a helpful AI a... | {"backend": "litellm", "model_name": "hosted_vllm/TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft", "n_responses": 1} | [
true
] | [
"B"
] | {"all_spans_summary": [{"answer": "B", "context": "Answer: B", "end_line": 1, "start_line": 1}], "empty_response": false, "extraction_method": "llm_judge", "final_span_info": {"answer": "B", "context": "Answer: B", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}, "is_final_of_multiple": false, "judge_mode... | [{"answer_block": null, "comparison_type": null, "correct_answer": null, "error": null, "evaluation_method": null, "extracted_answer": null, "final_answer": null, "is_correct": null, "is_equivalent": null, "method": "llm_judge", "reason": null}] | [
[
true
]
] | [
[
"B"
]
] | {"empty_response": false, "extraction_method": "llm_judge", "internal_spans_detailed": [{"answer": "B", "context": "Answer: B", "end_line": 1, "end_pos": 9, "start_line": 1, "start_pos": 0}], "is_final_of_multiple": false, "judge_model": "gpt-4.1-mini", "question_context": true, "span_positions": [{"end_pos": 9, "line_... | [{"value": "[{'answer_block': None, 'comparison_type': None, 'correct_answer': None, 'error': None, 'evaluation_method': None, 'extracted_answer': None, 'final_answer': None, 'is_correct': None, 'is_equivalent': None, 'method': 'llm_judge'}]"}] | 2025-07-28 |
End of preview. Expand in Data Studio
D-EVAL__standard_eval_v3__sft_gs__baseline_bon_tuning_n1-eval_sft
This evaluation dataset was created as part of the sft_gs__baseline_bon_tuning_n1 experiment using the SkillFactory experiment management system.
Evaluation Details
- Model Evaluated: TAUR-dev/M-sft_gs__baseline_bon_tuning_n1-sft
- Stage Name: eval_sft
- Experiment: sft_gs__baseline_bon_tuning_n1
- Total Samples: 2,450
Configuration
- Tasks: commonsenseQA, gsm8k, countdown_3arg, countdown_2arg, countdown_4arg, countdown_5arg, countdown_6arg, longmult_2dig, longmult_3dig, longmult_4dig, longmult_5dig
- Annotators: greedy
- Splits: test
Results Breakdown
By Task
- unknown: 2,450 samples
By Split
- test: 2,450 samples
Experiment Tracking
🔗 View complete experiment details: Experiment Tracker Dataset
The experiment tracker contains:
- Complete training logs and configurations
- Model lineage and relationships
- All evaluation results across stages
- Hyperparameter details and performance metrics
Dataset Usage
from datasets import load_dataset
# Load the complete evaluation dataset
dataset = load_dataset("TAUR-dev/D-EVAL__standard_eval_v3__sft_gs__baseline_bon_tuning_n1-eval_sft")
# Access different splits
for split_name, split_data in dataset.items():
print(f"Split {split_name}: {len(split_data)} samples")
# Example: Load specific split
test_data = load_dataset("TAUR-dev/D-EVAL__standard_eval_v3__sft_gs__baseline_bon_tuning_n1-eval_sft", split="test")
Evaluation Pipeline
This dataset was generated using the SkillFactory standard evaluation pipeline with:
- GPU-Optimized Hosting: Automatic GPU configuration for model hosting
- Multi-Annotator Support: greedy annotation types
- Comprehensive Metrics: Accuracy, performance, and detailed analysis
- Automatic Upload: Direct integration with HuggingFace Hub
Data Schema
Each sample contains:
- Input/Output: Original problem and model response
- Annotations: Evaluation results from different annotators
- Metadata: Task information, timing, and configuration details
- Metrics: Performance scores and analysis
This dataset was automatically generated by the SkillFactory ML experiment management system. For questions or issues, refer to the experiment tracker dataset.
- Downloads last month
- 7