Awesome RL with Human Feedback / 2024 / arXiv

Beyond One-Preference-Fits-All Alignment: Multi-Objective Direct Preference Optimization

作者：Zhanhui Zhou, Jie Liu, Chao Yang, Jing Shao, Yu Liu, Xiangyu Yue, Wanli Ouyang, Yu Qiao · 单位：Shanghai AI Laboratory, The Chinese University of Hong Kong · 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无奖励对齐
新标签打开 AlphaXiv新标签打开 Code
Beyond One-Preference-Fits-All Alignment: Multi-Objective Direct Preference Optimization figure
AlphaXiv 中文概览（可滚动查看）