Awesome RL with Human Feedback / 2020 and before / arXiv

Scalable agent alignment via reward modeling: a research direction

作者：Jan Leike, David Krueger, Tom Everitt, Miljan Martic, Vishal Maini, Shane Legg · 来源：arXiv · 发布日期：2020

奖励建模对齐目标设计策略梯度人类反馈强化学习样本效率
新标签打开 AlphaXiv新标签打开 Code
Scalable agent alignment via reward modeling: a research direction figure
AlphaXiv 中文概览（可滚动查看）