Awesome RL with Human Feedback / 2026 / arXiv

Text2Grad: Reinforcement Learning from Natural Language Feedback

作者：Hanyang Wang, Lu Wang, Chaoyun Zhang, Tianjun Mao, Si Qin, Qingwei Lin, Saravan Rajmohan, Dongmei Zhang · 单位：University of Chicago, Fudan University · 来源：arXiv · 发布日期：2025-05-28

奖励建模对齐目标设计人类反馈设计
新标签打开 AlphaXiv
Text2Grad: Reinforcement Learning from Natural Language Feedback figure
AlphaXiv 中文概览（可滚动查看）