Popular GitHub Repositories for Beaver

Safe RLHF: Constrained Value Alignment via Safe Reinforcement Learning from Human Feedback

BeaverTails is a collection of datasets designed to facilitate research on safety alignment in large language models (LLMs).