stratum-research/orthoqa-300

OrthoQA-300是一个结构化的、合成的数据集，包含300个关于骨科手术的患者-医生风格的问答对。该数据集模拟了真实的临床互动，其中问题以患者风格提出，答案则由语言模型生成，呈现医生风格。数据集按手术类型（如ACL重建、全髋关节置换）和主题（如“是什么？”，“恢复期”，“风险”）进行分组。它旨在作为早期阶段临床LLM研究的沙盒数据集，用于比较不同LLM的性能、测试判断性提示和幻觉检测管道、原型问答系统以及开发和测试评分系统。

OrthoQA-300 is a structured, synthetic dataset containing 300 patient-provider style question-and-answer pairs focused on orthopedic surgery. Each entry simulates a realistic clinical interaction with patient-style questions and LLM-generated provider-style answers. The dataset is grouped by surgical procedures (e.g., ACL Reconstruction, Total Hip Replacement) and themes (e.g., What is it?, Recovery, Risks). It is designed as a sandbox dataset for early-stage clinical LLM research, used for comparing different LLMs, testing judgment prompts and hallucination detection pipelines, prototyping question-answering systems, and developing and testing scoring systems.