Awesome RL with Human Feedback / 2024 / arXiv

Learning Reward and Policy Jointly from Demonstration and Preference Improves Alignment

作者：Chenliang Li, Siliang Zeng, Zeyi Liao, Jiaxiang Li, Dongyeop Kang, Alfredo Garcia, Mingyi Hong · 来源：arXiv · 发布日期：2024

奖励建模偏好学习对齐目标设计策略优化样本效率
新标签打开 AlphaXiv
Learning Reward and Policy Jointly from Demonstration and Preference Improves Alignment figure
AlphaXiv 中文概览（可滚动查看）