agentlans/claude-reasoning

---
license: apache-2.0
task_categories:
- text-generation
- question-answering
language:
- en
tags:
- reasoning
- synthetic
- claude
- chain-of-thought
- math
- code
- logic
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: thought
dtype: string
- name: source
dtype: string
---

# Claude Reasoning Dataset

This dataset is a curated collection of prompts and responses generated by **Claude by Anthropic**. It combines high-quality long reasoning data from multiple sources to provide a focused training set for models requiring logic, math, and coding capabilities.

If multiple answers were generated for the same input during the data collection process, the entry with the shortest reasoning content was selected to ensure conciseness and high signal-to-noise ratio.

## Dataset Details

### Composition

| Source | Rows |
|:---|---:|
| [Roman1111111/claude-sonnet-4.6-120000x](https://huggingface.co/datasets/Roman1111111/claude-sonnet-4.6-120000x) | 64 679 |
| [Roman1111111/claude-sonnet-4.6-100000X-filtered](https://huggingface.co/datasets/Roman1111111/claude-sonnet-4.6-100000X-filtered) | 51 973 |
| [Roman1111111/claude-opus-4.6-10000x](https://huggingface.co/datasets/Roman1111111/claude-opus-4.6-10000x) | 7 741 |
| [lordx64/reasoning-distill-claude-opus-4-7-max](https://huggingface.co/datasets/lordx64/reasoning-distill-claude-opus-4-7-max) | 4 592 |
| [TeichAI/claude-haiku-4.5-high-reasoning-1700x](https://huggingface.co/datasets/TeichAI/claude-haiku-4.5-high-reasoning-1700x) | 1 170 |
| [Hastagaras/Claude-Sonnet-4.6-Reasoning-Reddit-small](https://huggingface.co/datasets/Hastagaras/Claude-Sonnet-4.6-Reasoning-Reddit-small) | 480 |
| [Hastagaras/Claude-Sonnet-X-Opus-4.6-Reasoning-small-500](https://huggingface.co/datasets/Hastagaras/Claude-Sonnet-X-Opus-4.6-Reasoning-small-500) | 476 |
| [Hastagaras/Claude-Opus-4.6-Reasoning-BugFinder-400](https://huggingface.co/datasets/Hastagaras/Claude-Opus-4.6-Reasoning-BugFinder-400) | 411 |
| [Crownelius/Opus-4.6-Reasoning-3300x](https://huggingface.co/datasets/Crownelius/Opus-4.6-Reasoning-3300x) | 314 |
| [TeichAI/claude-sonnet-4.5-high-reasoning-250x](https://huggingface.co/datasets/TeichAI/claude-sonnet-4.5-high-reasoning-250x) | 77 |
| [TeichAI/claude-4.5-opus-high-reasoning-250x](https://huggingface.co/datasets/TeichAI/claude-4.5-opus-high-reasoning-250x) | 47 |

### Data Structure
Each row in the dataset follows this schema:
* **`question`**: The input prompt or problem statement.
* **`answer`**: The final response generated by Claude.
* **`thought`**: The reasoning chain produced by the model during inference.
* **`source`**: The origin dataset for the record.

**Example:**
```json
{
"question": "Matt can paint a house in 12 hours. Patty can paint the same house in one third the time. Rachel can paint the same house in 5 more than double the amount of hours as Patty. How long will it take Rachel to paint the house?",
"answer": "Patty can paint the house in 12 / 3 = 4 hours. Double Patty's time is 4 × 2 = 8 hours. Five more than that is 8 + 5 = 13 hours. It will take Rachel 13 hours to paint the house.",
"thought": "Simple math problem.\n\nMatt: 12 hours\nPatty: 12/3 = 4 hours\nRachel: 2*4 + 5 = 13 hours",
"source": "Roman1111111/claude-opus-4.6-10000x"
}
```

## Limitations

Despite being generated by state-of-the-art models, the outputs may still contain factual inaccuracies or reasoning errors.
Users are encouraged to verify the outputs for mission-critical applications.

## License

Apache 2.0