2 papers
cs.CL2025
ENCORE: Entropy-guided Reward Composition for Multi-head Safety Reward Models
Xiaomin Li, Xupeng Chen, Jingxuan Fan +2
The safety alignment of large language models (LLMs) often relies on reinforcement learning from human feedback (RLHF), which requires human annotations to construct preference dat…
cs.CL2025
Data-adaptive Safety Rules for Training Reward Models
Xiaomin Li, Mingye Gao, Zhiwei Zhang +2
Reinforcement Learning from Human Feedback (RLHF) is commonly employed to tailor models to human preferences, especially to improve the safety of outputs from large language models…