STAR Group

Safe & Trustworthy AI Resarch (STAR) group studies AI safety and trustworthiness. We are part of the State Key Laboratory of AI Safety.

At STAR Group, we believe interpretability is a key to building safe and trustworthy AI. Our research focuses on the knowledge mechanisms of AI models—how they learn, memorize, recall, update/edit, and forget knowledge. We also explore the security and privacy impacts of deploying AI in real-world applications, with a special focus on LLMs and recommender systems.