Representation-Aware Advantage Estimation: Your Reward Model Provides More Than A Scalar Output 文章

ArXiv CS.CL2026-06-10NEWSen作者: Guozheng Li, Xiyan Fu, Yiwen Guo

详细信息

来源站点
ArXiv CS.CL
作者
Guozheng Li, Xiyan Fu, Yiwen Guo
文章类型
NEWS
语言
en
发布日期
2026-06-10

摘要

arXiv:2606.10528v1 Announce Type: cross Abstract: Current reinforcement learning from human feedback (RLHF) methods primarily rely on scalar rewards from a trained reward model (RM). While effective, scalar rewards are often noisy and fail to capture fine-grained preference differences, whereas RM hidden states encode richer semantic and preference information. We introduce the representation-aware advantage estimation, which leverages RM hidden states and models them as auxiliary signals for better advantage estimation. Specifically, we propose the Graph-based Advantage Estimation (GraphAE), treat each sampled group as a graph, where nodes correspond to responses and edges capture their similarity in the RM hidden space. Then advantages are computed via graph propagation, enabling each sample to incorporate contextual information from its neighbors. GraphAE is lightweight and can be seamlessly integrated into existing group-based RL algorithms.

相关事件

暂无数据

相关公司

暂无数据

相关人物

暂无数据

相关产品

暂无数据