This dataset is a mathematical problem reasoning dataset, released in 2025 by a research team from Stanford University and the University of Washington. It aims to strengthen the logical coherence of large language models (LLMs) and optimize their structured thinking through mathematical reasoning. The relevant paper results are:s1: Simple test-time scaling".

This dataset contains 1,000 samples, focusing on mathematical problems and reasoning trajectories, covering multiple mathematical fields such as algebra, geometry, and probability. Each sample includes a problem description, solution steps, answer, and reasoning trajectory generated by DeepSeek r1.