TY  - RPRT
TI  - Evaluating Reward Model Generalization via Pairwise Maximum Discrepancy Competitions
AU  - Shunyang Luo
AU  - Peibei Cao
AU  - Zhihui Zhu
AU  - Kehua Feng
AU  - Zhihua Wang
AU  - Keyan Ding
PY  - 2026
UR  - https://arxiv.org/abs/2601.16987
ID  - 2601.16987
ER  -