TY  - RPRT
TI  - Hybrid Reward Normalization for Process-supervised Non-verifiable Agentic Tasks
AU  - Peiran Xu
AU  - Zhuohao Li
AU  - Xiaoying Xing
AU  - Guannan Zhang
AU  - Debiao Li
AU  - Kunyu Shi
PY  - 2025
UR  - https://arxiv.org/abs/2509.25598
ID  - 2509.25598
ER  -