openmed-community/MedReason-Stenographic

该数据集包含31,535个医学问答对，带有速记推理痕迹，使用MiniMax M2.1从原始UCSC-VLAA/MedReason数据集生成。数据集将医学问答推理转换为速记格式，使用符号协议设计，以实现高密度、机器可解析的推理痕迹。这种格式消除了自然语言填充，同时保持了医学推理的完整逻辑流程。关键特征包括31,535个带有压缩推理的医学问答样本、使用符号表示的速记推理痕迹、在thinking字段中保留的完整思维过程，以及基于获奖的MedReason数据集（2025年HuggingFace推理数据集竞赛第三名）。

This dataset contains 31,535 medical question-answer pairs with stenographic reasoning traces, generated using MiniMax M2.1 from the original UCSC-VLAA/MedReason dataset. The dataset transforms medical QA reasoning into a stenographic format using a symbolic protocol designed for high-density, machine-parseable reasoning traces. This format eliminates natural language filler while maintaining the complete logical flow of medical reasoning. Key features include 31,535 samples of medical QA with compressed reasoning (filtered for non-empty reasoning), stenographic reasoning traces using symbolic notation, full thinking process preserved in the thinking field, and based on the award-winning MedReason dataset (3rd place, HuggingFace Reasoning Datasets Competition 2025).