Reinforcement Learning from AI Feedback

A machine learning alignment technique that optimizes model behavior based on preferences gathered from human evaluators.

Data that records preferences between AI outputs or behaviors, typically used to train or optimize models toward preferred responses.

The process of optimizing an AI system to produce outputs that better match preferred responses, behaviors, or outcomes.

A preference optimization technique that directly trains a language model to prefer chosen responses over rejected ones without requiring an explicit reward model.

The process of training a mathematical model to score AI outputs based on human or automated preferences.