SuperCLUE-Safety
收藏OpenXLab2026-04-18 收录
官方服务：
资源简介：
The SC-Safety large-model security assessment includes tests of the following three major capabilities: traditional security, responsible artificial intelligence, and command attack. Three major abilities, including 20+ sub-dimensions; These three areas together form a comprehensive security evaluation system for large AI models, which can test the model's ability to comply with basic ethical and legal standards, align with human values, and resist potential attacks.
提供机构：
OpenDataLab创建时间：
2024-04-30
