The Alignment Ceiling: Objective Mismatch in Reinforcement Learning from Human Feedback

Oct 31, 2023
|
Nathan Lambert, Roberto Calandra
Tags:
Reinforcement Learning
Download the paper!Read on ArXiv!Run the code!Video available!
ABSTRACT:
hide & show ↓↑

What you need to know:

Citation

‍

Links

Download the paper!
Read on Arxiv!
Read the code!

Video

←All ResearchGoogle Scholar

Tags

Novel Robots
Reinforcement Learning
Multi-Agent Systems
Beneficial AI
Robotics
Model Learning