The Alignment Ceiling: Objective Mismatch in Reinforcement Learning from Human Feedback
CoRR（2023）
Key words
human feedback,reinforcement learning,objective mismatch,alignment
AI Read Science
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined