Awesome RL with Human Feedback / 2023 / arXiv

Aligning Text-to-Image Models using Human Feedback

作者：Kimin Lee, Hao liu, MoonKyung Ryu, Olivia Watkins, Yuqing Du, Craig Boutilier, Pieter Abbeel, Mohammad Ghavamzadeh, Shixiang Shane Gu · 来源：arXiv · 发布日期：2023

奖励建模偏好学习对齐目标设计训练稳定性多模态反馈
新标签打开 AlphaXiv
Aligning Text-to-Image Models using Human Feedback figure
AlphaXiv 中文概览（可滚动查看）