Fetching the paper…

PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference · Around