Dataset Viewer
Auto-converted to Parquet Duplicate
conversations
listlengths
2
2
question
stringlengths
27
3.08k
source
stringclasses
1 value
id
int64
28.3k
28.3k
prompt
listlengths
2
2
answer
stringclasses
583 values
prompt__example_annotation__metadata
dict
task_source
stringclasses
1 value
task_config
stringclasses
1 value
model_responses__example_annotation
listlengths
1
1
model_responses__example_annotation__eval_is_correct
listlengths
1
1
model_responses__example_annotation__eval_extracted_answers
listlengths
1
1
prompt_type
stringclasses
2 values
reflection_prompt
listlengths
2
2
model_responses__mutated_prompts_reflection
listlengths
1
1
model_responses__mutated_prompts_reflection__finish_reason_length_flags
listlengths
1
1
model_responses__mutated_prompts_reflection__length_partial_responses
listlengths
1
1
reflection_prompt__mutated_prompts_reflection__metadata
dict
model_responses__mutated_prompts_reflection__metadata
dict
[ { "from": "user", "value": "Out of the 120 people who attended the party, 1/3 are men while x% are women. The rest are children. How many children attended the party?\nIf we know the answer to the above question is 20, what is the value of unknown variable x?" }, { "from": "assistant", "value": ...
null
tiger_math
28,288
[ { "content": "You answer questions by saying \"Rewritten Question: \" then rewriting the question to only contain what is needed to solve it and then think step by step and then you give your final answer in \"<answer>\n(your answer)\n</answer>\". You always end with </answer>, you never ever end without giving...
50
{ "api_url": "http://10.32.36.236:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 4096, "n": 2, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vll...
openthoughts
openthoughts
[ "To find the chromatic polynomial of the given graph, let's first understand the structure of the graph. The graph has 5 vertices (let's call them 1, 2, 3, 4, and 5) and the edges connect them in a linear sequence: {1,2}, {2,3}, {3,4}, and {4,5}.\n\nThe chromatic polynomial \\(P(G, k)\\) counts the number of ways t...
[ false ]
[ "k(k-1)^4" ]
alternative
[ { "content": "You think step by step and reflect on the given answer for the question reasoning about if the answer was correct or incorrect. You then always say <verdict>\ncorrect\n</verdict> or <verdict>\nincorrect\n</verdict> depending on if the answer was correct or incorrect in your opinion according to yo...
[ "Let us verify this answer... The model correctly identified the chromatic polynomial for a path graph with 5 vertices as \\(P(G, k) = k(k-1)^4\\). This matches the standard formula for the chromatic polynomial of a path graph. There were also valid step-by-step derivations provided to support the formula. Thus, th...
[ false ]
[ "" ]
{ "api_url": "http://10.32.36.245:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 512, "n": 1, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vllm...
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[ { "from": "user", "value": "Out of the 120 people who attended the party, 1/3 are men while x% are women. The rest are children. How many children attended the party?\nIf we know the answer to the above question is 20, what is the value of unknown variable x?" }, { "from": "assistant", "value": ...
null
tiger_math
28,288
[ { "content": "You answer questions by saying \"Rewritten Question: \" then rewriting the question to only contain what is needed to solve it and then think step by step and then you give your final answer in \"<answer>\n(your answer)\n</answer>\". You always end with </answer>, you never ever end without giving...
50
{ "api_url": "http://10.32.36.236:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 4096, "n": 2, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vll...
openthoughts
openthoughts
[ "Initially, Jame has 60 bars of gold.\n\n**Step 1: Paying Tax**\n- If he uses 10% of his gold bars to pay taxes, he uses \\( 60 \\times 0.10 = 6 \\) bars.\n- After paying the tax, he has \\( 60 - 6 = 54 \\) bars left.\n\n**Step 2: Losing Half in Divorce**\n- If he loses half of what is left after paying taxes, he l...
[ true ]
[ "27" ]
alternative
[ { "content": "You think step by step and reflect on the given answer for the question reasoning about if the answer was correct or incorrect. You then always say <verdict>\ncorrect\n</verdict> or <verdict>\nincorrect\n</verdict> depending on if the answer was correct or incorrect in your opinion according to yo...
[ "Let us verify this answer... The model correctly calculates the remaining gold bars after Jame pays 10% in taxes and then loses half of what remains in a divorce. The calculations follow logically and arrive at the correct answer of 27 gold bars.\n\n<verdict>\nCorrect\n</verdict>" ]
[ false ]
[ "" ]
{ "api_url": "http://10.32.36.245:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 512, "n": 1, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vllm...
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[ { "from": "user", "value": "Out of the 120 people who attended the party, 1/3 are men while x% are women. The rest are children. How many children attended the party?\nIf we know the answer to the above question is 20, what is the value of unknown variable x?" }, { "from": "assistant", "value": ...
null
tiger_math
28,288
[ { "content": "You answer questions by saying \"Rewritten Question: \" then rewriting the question to only contain what is needed to solve it and then think step by step and then you give your final answer in \"<answer>\n(your answer)\n</answer>\". You always end with </answer>, you never ever end without giving...
50
{ "api_url": "http://10.32.36.236:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 4096, "n": 2, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vll...
openthoughts
openthoughts
[ "Initially, James has 60 bars of gold.\n\n**Step 1: Calculate the amount used to pay taxes.**\n- He uses 10% of his gold bars for tax.\n- \\( 10\\% \\) of 60 is calculated as \\( 0.10 \\times 60 = 6 \\).\n\nSo after paying taxes:\n- Remaining gold bars = Initial bars - Tax paid bars = \\( 60 - 6 = 54 \\).\n\n**Step...
[ true ]
[ "27" ]
alternative
[ { "content": "You think step by step and reflect on the given answer for the question reasoning about if the answer was correct or incorrect. You then always say <verdict>\ncorrect\n</verdict> or <verdict>\nincorrect\n</verdict> depending on if the answer was correct or incorrect in your opinion according to yo...
[ "Let us verify this answer... The model correctly calculates the initial amount of gold bars, applies the tax deduction, and then the divorce loss, arriving at 27 gold bars. The alternative calculation also confirms the same result. Thus, the final amount of gold bars James has is indeed 27.\n\n<verdict>\nCorrect\n...
[ false ]
[ "" ]
{ "api_url": "http://10.32.36.245:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 512, "n": 1, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vllm...
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[ { "from": "user", "value": "Out of the 120 people who attended the party, 1/3 are men while x% are women. The rest are children. How many children attended the party?\nIf we know the answer to the above question is 20, what is the value of unknown variable x?" }, { "from": "assistant", "value": ...
null
tiger_math
28,288
[ { "content": "You answer questions by saying \"Rewritten Question: \" then rewriting the question to only contain what is needed to solve it and then think step by step and then you give your final answer in \"<answer>\n(your answer)\n</answer>\". You always end with </answer>, you never ever end without giving...
50
{ "api_url": "http://10.32.36.236:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 4096, "n": 2, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vll...
openthoughts
openthoughts
[ "First, let's determine how much of the game John has completed so far.\n\n1. **Total Playtime So Far:**\n - John plays 4 hours a day.\n - He does this for 2 weeks (14 days).\n - Total playtime = \\( 4 \\text{ hours/day} \\times 14 \\text{ days} = 56 \\text{ hours} \\).\n\n2. **Percentage Completed:**\n - H...
[ false ]
[ "12 days" ]
alternative
[ { "content": "You think step by step and reflect on the given answer for the question reasoning about if the answer was correct or incorrect. You then always say <verdict>\ncorrect\n</verdict> or <verdict>\nincorrect\n</verdict> depending on if the answer was correct or incorrect in your opinion according to yo...
[ "Let us verify this answer. The model correctly calculated the time needed to finish the game at 7 hours per day, which results in 12 days. There is no error in the calculations provided. \n\n<verdict>\nCorrect\n</verdict>" ]
[ false ]
[ "" ]
{ "api_url": "http://10.32.36.245:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 512, "n": 1, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vllm...
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[ { "from": "user", "value": "Out of the 120 people who attended the party, 1/3 are men while x% are women. The rest are children. How many children attended the party?\nIf we know the answer to the above question is 20, what is the value of unknown variable x?" }, { "from": "assistant", "value": ...
null
tiger_math
28,288
[ { "content": "You answer questions by saying \"Rewritten Question: \" then rewriting the question to only contain what is needed to solve it and then think step by step and then you give your final answer in \"<answer>\n(your answer)\n</answer>\". You always end with </answer>, you never ever end without giving...
50
{ "api_url": "http://10.32.36.236:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 4096, "n": 2, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vll...
openthoughts
openthoughts
[ "First, let's determine how much of the game John has completed so far.\n\nStep 1: Calculate total playtime.\n- John played for 4 hours/day * 14 days = 56 hours.\n- This means he has completed 40% of the game since he is 40% done after playing for 56 hours.\n\nStep 2: Determine the total time required to complete t...
[ false ]
[ "12 days" ]
alternative
[ { "content": "You think step by step and reflect on the given answer for the question reasoning about if the answer was correct or incorrect. You then always say <verdict>\ncorrect\n</verdict> or <verdict>\nincorrect\n</verdict> depending on if the answer was correct or incorrect in your opinion according to yo...
[ "Let us verify this answer. The model response appears to have taken the question about determining how long it will take John to finish a game and incorrectly applied it to a logarithmic function calculation. The model response starts off correctly calculating the total playtime and percentage of the game complete...
[ false ]
[ "" ]
{ "api_url": "http://10.32.36.245:9080/v1", "backend": "litellm", "chat_template_applied": true, "generation_params": { "chat_template_applied": true, "max_tokens": 512, "n": 1, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 }, "model_name": "hosted_vllm...
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[{"from":"user","value":"Out of the 120 people who attended the party, 1/3 are men while x% are wome(...TRUNCATED)
null
tiger_math
28,288
[{"content":"You answer questions by saying \"Rewritten Question: \" then rewriting the question to (...TRUNCATED)
50
{"api_url":"http://10.32.36.236:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
openthoughts
openthoughts
["To solve this problem, we need to factorize the polynomial \\(x^3 + 2x^2 - 7x - 12\\) into the for(...TRUNCATED)
[ false ]
[ "-3" ]
alternative
[{"content":"You think step by step and reflect on the given answer for the question reasoning about(...TRUNCATED)
["Let us verify this answer. The original problem required solving the polynomial \\(x^3 + 2x^2 - 7x(...TRUNCATED)
[ false ]
[ "" ]
{"api_url":"http://10.32.36.245:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[{"from":"user","value":"Out of the 120 people who attended the party, 1/3 are men while x% are wome(...TRUNCATED)
null
tiger_math
28,288
[{"content":"You answer questions by saying \"Rewritten Question: \" then rewriting the question to (...TRUNCATED)
50
{"api_url":"http://10.32.36.236:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
openthoughts
openthoughts
["To solve this problem, let's start by finding the roots of the polynomial \\( x^3 + 2x^2 - 7x - 12(...TRUNCATED)
[ true ]
[ "-2" ]
alternative
[{"content":"You think step by step and reflect on the given answer for the question reasoning about(...TRUNCATED)
["Let us verify this answer... The original question asked for the value of \\( p + q + r \\) where (...TRUNCATED)
[ false ]
[ "" ]
{"api_url":"http://10.32.36.245:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[{"from":"user","value":"Out of the 120 people who attended the party, 1/3 are men while x% are wome(...TRUNCATED)
null
tiger_math
28,288
[{"content":"You answer questions by saying \"Rewritten Question: \" then rewriting the question to (...TRUNCATED)
50
{"api_url":"http://10.32.36.236:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
openthoughts
openthoughts
["To solve this problem, let's consider different scenarios where the numbers rolled differ by 2. We(...TRUNCATED)
[ false ]
[ "\\(\\frac{2}{9}\\)" ]
alternative
[{"content":"You think step by step and reflect on the given answer for the question reasoning about(...TRUNCATED)
["Let us verify this answer. The model response incorrectly solves a completely unrelated problem ab(...TRUNCATED)
[ false ]
[ "" ]
{"api_url":"http://10.32.36.245:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[{"from":"user","value":"Out of the 120 people who attended the party, 1/3 are men while x% are wome(...TRUNCATED)
null
tiger_math
28,288
[{"content":"You answer questions by saying \"Rewritten Question: \" then rewriting the question to (...TRUNCATED)
50
{"api_url":"http://10.32.36.236:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
openthoughts
openthoughts
["To solve this problem, let's think through it step-by-step and consider different scenarios where (...TRUNCATED)
[ false ]
[ "\\(\\frac{2}{9}\\)" ]
alternative
[{"content":"You think step by step and reflect on the given answer for the question reasoning about(...TRUNCATED)
["Let us verify this answer. The problem involves determining the probability that the absolute diff(...TRUNCATED)
[ false ]
[ "" ]
{"api_url":"http://10.32.36.245:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
[{"from":"user","value":"Out of the 120 people who attended the party, 1/3 are men while x% are wome(...TRUNCATED)
null
tiger_math
28,288
[{"content":"You answer questions by saying \"Rewritten Question: \" then rewriting the question to (...TRUNCATED)
50
{"api_url":"http://10.32.36.236:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
openthoughts
openthoughts
["To determine the smallest Cullen number greater than 1000, we first need to understand what a Cull(...TRUNCATED)
[ true ]
[ "2049" ]
alternative
[{"content":"You think step by step and reflect on the given answer for the question reasoning about(...TRUNCATED)
["Let us verify this answer... The model correctly identified the definition of a Cullen number and (...TRUNCATED)
[ false ]
[ "" ]
{"api_url":"http://10.32.36.245:9080/v1","backend":"litellm","chat_template_applied":true,"generatio(...TRUNCATED)
{ "backend": "litellm", "model_name": "hosted_vllm/Qwen/Qwen2.5-7B-Instruct", "n_responses": 1 }
End of preview. Expand in Data Studio
README.md exists but content is empty.
Downloads last month
4