This is a filtered and metadata enriched version of open-thoughts/OpenThoughts-114k. While the original dataset is a valuable resource containing DeepSeek-R1 outputs, it has very little metadata (only 2 fields: system and conversations). It does not contain, for instance, the original solution label, which means that we can not verify the model answers.

  What we did

filtered the dataset for math content (math questions were prefixed by "Return your final response within… See the full description on the dataset page: https://huggingface.co/datasets/open-r1/OpenThoughts-114k-math.

gadkins/open-thoughts-math-dry-run dataset hosted on Hugging Face and contributed by the HF Datasets community

mlfoundations-dev/openthoughts-114k-no-special-template_eval_03-11-25_05-44-46_f912

Precomputed model outputs for evaluation.

  Evaluation Results





  GPQADiamond

Average Accuracy: 31.31% ± 4.97% Number of Runs: 3

Run Accuracy Questions Solved Total Questions

1 24.24% 48 198

2 26.26% 52 198

3 43.43% 86 198

GeneralThought-195K

NEWEST RELEASE WITH 323K TRACES IS HERE

Thought wants to be free

Open reasoning data from the General Reasoning resource for March 3 2025. The dataset contains questions, reference answers, reasoning traces, final answers and other metadata from several popular reasoning models including DeepSeek-R1, DeepSeek-R1-Zero, OpenThoughts-32B, LIMO, deepseek-r1-distill-llama-70b, DeepHermes-3-Llama-3-8B-Preview and DeepScaleR-1.5B-Preview. We also include final… See the full description on the dataset page: https://huggingface.co/datasets/GeneralReasoning/GeneralThought-195K.

mlfoundations-dev/DCFT-open-thoughts-subset-v1-etash_1742823125_eval_0981

Precomputed model outputs for evaluation.

  Evaluation Results





  Summary

Metric AIME24 AIME25 AMC23 MATH500 GPQADiamond LiveCodeBench

Accuracy 22.7 16.7 63.5 80.4 22.4 31.5

  AIME24

Average Accuracy: 22.67% ± 1.46% Number of Runs: 5

Run Accuracy Questions Solved Total Questions

1 23.33% 7 30

2 23.33% 7 30

3 23.33% 7 30

4 26.67% 8 30

5 16.67% 5 30… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-dev/DCFT-open-thoughts-subset-v1-etash_1742823125_eval_0981.

mlfoundations-dev/DCFT-open-thoughts-subset-claude-v1-etash_1742633651_eval_0981

Precomputed model outputs for evaluation.

  Evaluation Results





  Summary

Metric AIME24 AIME25 AMC23 MATH500 GPQADiamond LiveCodeBench

Accuracy 22.0 22.0 61.0 81.8 23.6 31.4

  AIME24

Average Accuracy: 22.00% ± 2.02% Number of Runs: 5

Run Accuracy Questions Solved Total Questions

1 23.33% 7 30

2 20.00% 6 30

3 20.00% 6 30

4 30.00% 9 30

5… See the full description on the dataset page: https://huggingface.co/datasets/mlfoundations-dev/DCFT-open-thoughts-subset-claude-v1-etash_1742633651_eval_0981.

GeneralThought-430K

Thought wants to be free

Open reasoning data from the General Reasoning resource for March 14 2025. The dataset contains questions, reference answers, reasoning traces, final answers and other metadata from several popular reasoning models including DeepSeek-R1, DeepSeek-R1-Zero, OpenThoughts-32B, LIMO, deepseek-r1-distill-llama-70b, DeepHermes-3-Llama-3-8B-Previewand DeepScaleR-1.5B-Preview. We also include final answers from o3-mini-2025-01-31… See the full description on the dataset page: https://huggingface.co/datasets/GeneralReasoning/GeneralThought-430K.