TY  - RPRT
TI  - Information-Theoretic Reward Decomposition for Generalizable RLHF
AU  - Liyuan Mao
AU  - Haoran Xu
AU  - Amy Zhang
AU  - Weinan Zhang
AU  - Chenjia Bai
PY  - 2025
UR  - https://arxiv.org/abs/2504.06020
ID  - 2504.06020
ER  -