Beijing Institute of AI Safety and Governance

AI & ML interests

AI Safety, AI Ethics, AI Governance, Risk Mitigation

Recent Activity

floyed  authored a paper 2 days ago

Safety Instincts: LLMs Learn to Trust Their Internal Compass for Self-Defense

floyed  authored a paper 2 days ago

VESPO: Variational Sequence-Level Soft Policy Optimization for Stable Off-Policy LLM Training

floyed  authored a paper 2 days ago

StressPrompt: Does Stress Impact Large Language Models and Human Performance Similarly?

View all activity

Team members 12