Glossary · Model Training

Reinforcement Learning

A training paradigm where a model learns by taking actions and receiving reward signals, used in RLHF-style alignment training and in some agent-planning research.