GeneralThought-Feb25

# GeneralThought-Feb25

> Thought wants to be free

**NEWEST RELEASE WITH 195K TRACES IS [HERE](https://huggingface.co/datasets/GeneralReasoning/GeneralThought-195K)**

------

Open reasoning data from the [General Reasoning](https://gr.inc) resource for February 2025.

The dataset contains questions, reference answers, reasoning traces, final answers and other metadata from several popular reasoning models including [DeepSeek-R1](https://gr.inc/DeepSeek/models/DeepSeek-R1/), [DeepSeek-R1-Zero](https://gr.inc/DeepSeek/models/DeepSeek-R1-Zero/), [OpenThoughts-32B](https://gr.inc/open-thoughts/models/OpenThinker-32B/), [LIMO](https://gr.inc/GAIR-NLP/models/LIMO/), [deepseek-r1-distill-llama-70b](https://gr.inc/DeepSeek/models/deepseek-r1-distill-llama-70b/), [DeepHermes-3-Llama-3-8B-Preview](https://gr.inc/NousResearch/models/DeepHermes-3-Llama-3-8B-Preview/) and [DeepScaleR-1.5B-Preview](https://gr.inc/agentica-org/models/DeepScaleR-1.5B-Preview/). We also include final answers from [o3-mini-2025-01-31](https://gr.inc/OpenAI/models/o3-mini-2025-01-31/) and [gemini-2-flash-thinking-exp-01-21](https://gr.inc/Google/models/gemini-2-flash-thinking-exp-01-21/) for comparison and evaluation. The February release has 123,394 rows of data.

## Metadata

A row of data contains the following information:

```python
row = {
'question_id': '296582', # question ID on the GR platform
'question_url': 'https://gr.inc/question/of-a-quadrilateral-if-its-angle-measures-are-in-the-ratio-of-3456-find-the-m', # URL on gr.inc
'question': 'Of a quadrilateral if its angle measures are in the ratio of 3:4:5:6, find the measure of each angle.' # Question text
'reference_answer': '60°, 80°, 100°, 120°', # Reference answer
'model_name': 'DeepSeek/DeepSeek-R1', # The model that generated the trace
'model_answer': 'The measures of the angles in the quadrilateral are calculated as follows...', # the model final answer text
'model_reasoning': 'Okay, so I need to find the measures of each angle in a quadrilateral...' # the model reasoning text
'task': 'Applying Ratios to Angle-Measure Sums', # name of the task on GR
'question_license': 'MIT', # license of the question
'question_source': 'General/VNet', # original dataset source or author on GR
'community_question_score': 0 # community score for the question on GR; negative means downvoted, positive upvoted,
'community_answer_score': 0, # community score for the answer on GR; negative means downvoted, positive upvoted
'verifier_score': 1.0 # an average verification score between 0-1; if multiple verifiers, this could be between, e.g. 0.5 if one verifier marks as correct, another incorrect
}
```

## How can I use the data?

The dataset is a great complement to [OpenThoughts-114k](https://huggingface.co/datasets/open-thoughts/OpenThoughts-114k), [OpenR1](https://huggingface.co/datasets/open-r1/OpenR1-Math-Raw), [SYNTHETIC-1](https://huggingface.co/datasets/PrimeIntellect/SYNTHETIC-1), among others!

Here's some example use cases for our dataset:

- Perform SFT distillation and use it to train a small reasoning model.
- Ablate alongside datasets from our open source friends (and see if cross-group diversity helps).
- Analyse reasoning differences between models: reasoning length, language switching, and use of connectors like "wait" and "alternatively".

The verification side of the GR platform is still early, so we would recommend focusing on distillation rather than RL for now.

## Thanks

Thank you to the contributors of questions for this dataset:

First - thanks to the questions we sourced from folks like [Numina](https://huggingface.co/datasets/AI-MO/NuminaMath-CoT), [SCP-116k](https://huggingface.co/datasets/EricLu/SCP-116K), [natural_reasoning](https://huggingface.co/datasets/facebook/natural_reasoning) and others! We've credited you in the question_source field of each row of the dataset.

Thanks to GR community contributors who contributed:
- **Questions**: Tim_tom_0, Jarius and panpan.
- **Verifications**: knight_raider, supahao, otaldohenrikk, Jarius and panpan.
- **Quality Votes**: Doge, yuchen.zhang2003, eli5, pcpthm, Esac, tginart, arpitg1991, yych, panpan, Jarius, supahao and knight-raider.

Going forward we will continue to credit those who contribute questions in future data dumps on Hugging Face 🤗.

**We will look to publish a paper with co-authorship for contributors.**

# GeneralThought-Feb25

> 思维本应自由

**本次发布的包含19.5万条推理轨迹的最新版本已上线[此处](https://huggingface.co/datasets/GeneralReasoning/GeneralThought-195K)**

------

本数据集为2025年2月发布的[通用推理（General Reasoning）](https://gr.inc)开源推理数据资源。

数据集涵盖问题、参考答案、推理轨迹、最终答案及其他元数据（metadata），涉及的主流推理模型包括[DeepSeek-R1](https://gr.inc/DeepSeek/models/DeepSeek-R1/)、[DeepSeek-R1-Zero](https://gr.inc/DeepSeek/models/DeepSeek-R1-Zero/)、[OpenThoughts-32B](https://gr.inc/open-thoughts/models/OpenThinker-32B/)、[LIMO](https://gr.inc/GAIR-NLP/models/LIMO/)、[deepseek-r1-distill-llama-70b](https://gr.inc/DeepSeek/models/deepseek-r1-distill-llama-70b/)、[DeepHermes-3-Llama-3-8B-Preview](https://gr.inc/NousResearch/models/DeepHermes-3-Llama-3-8B-Preview/)以及[DeepScaleR-1.5B-Preview](https://gr.inc/agentica-org/models/DeepScaleR-1.5B-Preview/)。为便于对比与评估，本数据集还纳入了[o3-mini-2025-01-31](https://gr.inc/OpenAI/models/o3-mini-2025-01-31/)与[gemini-2-flash-thinking-exp-01-21](https://gr.inc/Google/models/gemini-2-flash-thinking-exp-01-21/)的最终答案。本次2月发布的数据集共包含123394条数据样本。

## 元数据

单条数据样本包含如下字段：

python
row = {
'question_id': '296582', # GR平台上的问题ID
'question_url': 'https://gr.inc/question/of-a-quadrilateral-if-its-angle-measures-are-in-the-ratio-of-3456-find-the-m', # gr.inc平台上的问题链接
'question': '已知某四边形的内角比例为3:4:5:6，求各内角的度数。', # 问题文本
'reference_answer': '60°, 80°, 100°, 120°', # 参考答案
'model_name': 'DeepSeek/DeepSeek-R1', # 生成推理轨迹的模型名称
'model_answer': '该四边形各内角的度数计算过程如下...', # 模型生成的最终答案文本
'model_reasoning': '好的，我需要计算该四边形各内角的度数...', # 模型生成的推理过程文本
'task': '比例应用于内角和计算', # GR平台上的任务名称
'question_license': 'MIT', # 问题的开源许可证
'question_source': 'General/VNet', # GR平台上的原始数据集来源或作者
'community_question_score': 0, # GR平台上的社区问题评分：负值代表差评，正值代表好评，0为无评分
'community_answer_score': 0, # GR平台上的社区答案评分：负值代表差评，正值代表好评，0为无评分
'verifier_score': 1.0 # 验证者平均评分，取值范围0-1；若存在多名验证者，评分可能为区间值，例如0.5代表一名验证者判定正确、另一名判定错误
}

## 如何使用本数据集？

本数据集可作为[OpenThoughts-114k](https://huggingface.co/datasets/open-thoughts/OpenThoughts-114k)、[OpenR1](https://huggingface.co/datasets/open-r1/OpenR1-Math-Raw)、[SYNTHETIC-1](https://huggingface.co/datasets/PrimeIntellect/SYNTHETIC-1)等多款开源数据集的优质补充！

本数据集的典型应用场景包括：
- 开展监督微调（Supervised Fine-Tuning, SFT）蒸馏，用于训练轻量化推理模型
- 与其他开源伙伴的数据集联合进行消融实验，探究跨群体多样性对模型性能的影响
- 分析不同模型间的推理差异：包括推理长度、语言切换行为，以及“等一下”“或者”等衔接词的使用习惯

GR平台的验证模块仍处于早期阶段，因此我们建议现阶段优先将本数据集用于蒸馏任务，而非强化学习（Reinforcement Learning, RL）相关训练。

## 致谢

谨为本数据集的问题贡献者致谢：

首先，感谢我们从[Numina](https://huggingface.co/datasets/AI-MO/NuminaMath-CoT)、[SCP-116k](https://huggingface.co/datasets/EricLu/SCP-116K)、[natural_reasoning](https://huggingface.co/datasets/facebook/natural_reasoning)等资源中获取的问题贡献者！我们已在每条数据样本的`question_source`字段中为您署名。

同时感谢GR社区的贡献者：
- **问题贡献**：Tim_tom_0、Jarius与panpan
- **验证工作**：knight_raider、supahao、otaldohenrikk、Jarius与panpan
- **质量评分**：Doge、yuchen.zhang2003、eli5、pcpthm、Esac、tginart、arpitg1991、yych、panpan、Jarius、supahao与knight-raider

未来我们将继续在Hugging Face 🤗发布的数据集版本中为问题贡献者署名。

**我们计划为贡献者提供联合署名发表学术论文的机会。**