Awesome RL with Human Feedback / 2022 / arXiv

Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback

作者：Yuntao Bai, Andy Jones, Kamal Ndousse, et al · 来源：arXiv · 发布日期：2022-04-12

安全对齐鲁棒性/噪声反馈偏好学习在线人类反馈强化学习人类反馈设计
新标签打开 AlphaXiv新标签打开 Code新标签打开 Dataset
Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback figure
AlphaXiv 中文概览（可滚动查看）