Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback

Stephen Casper*, Xander Davies*, …, Andi Peng, …, Dylan Hadfield-Menell

May 2023