GeneralThought-430K Large-Scale Reasoning Dataset

GeneralThought-430K is a large-scale reasoning dataset released by the General Reasoning team in 2025. It aims to provide standardized resources for training and evaluating the logical reasoning, interdisciplinary knowledge integration, and complex problem-solving capabilities of large language models.

The dataset contains 430,000 samples, covering problems in the fields of mathematics, code, physics, chemistry, natural sciences, humanities and social sciences, engineering technology, etc. It includes questions, reference answers, reasoning trajectories, final answers and other metadata from multiple reasoning models, including DeepSeek-R1, DeepSeek-R1-Zero, OpenThoughts-32B, LIMO and other mainstream models. The final answers of o3-mini-2025-01-31, gemini-2-flash-thinking-exp-01-21 and claude-3-7-sonnet-20250219 are also included for comparison and evaluation.