Awesome RL with Human Feedback / 2020 and before / arXiv

Fine-Tuning Language Models from Human Preferences

作者：Daniel M. Ziegler, Nisan Stiennon, Jeffrey Wu, Tom B. Brown, Alec Radford, Dario Amodei, Paul Christiano, Geoffrey Irving · 来源：arXiv · 发布日期：2019-09-18

奖励建模偏好学习策略优化长文本/对话
新标签打开 AlphaXiv新标签打开 Code新标签打开 Dataset
Fine-Tuning Language Models from Human Preferences figure
AlphaXiv 中文概览（可滚动查看）