OpenStellarTeam/Chinese-SafetyQA

Chinese SafetyQA是一个创新的基准，旨在评估大型语言模型在中文安全领域短篇事实性问题上的事实性能力。该数据集包含7个主要主题、27个次要主题和103个细分主题，涵盖了广泛的安全领域知识。数据集的特点包括中文语言、无害内容、多样性、易于评估、静态性和挑战性。此外，数据集还提供了两种格式的数据：短篇问答（QA）和多项选择题（MCQ），以便用户轻松测试模型的安全知识边界。

Chinese SafetyQA is an innovative benchmark designed to evaluate the factuality ability of large language models, specifically for short-form factual questions in the Chinese safety domain. The dataset includes 7 major topics, 27 secondary topics, and 103 diverse subtopics, covering a wide range of safety-related knowledge. Key features of the dataset include Chinese language, harmless content, diversity, easy-to-evaluate, static, and challenging. Additionally, the dataset provides data in two formats: short-form question-answer (QA) and multiple-choice questions (MCQ), allowing users to easily test the boundaries of a model’s safety knowledge.