An area of machine learning concerned with how intelligent agents ought to take actions in an environment to maximize cumulative reward.
Reinforcement Learning (RL) allows AI systems to learn complex behaviors through trial and error. It is the underlying mechanism for RLHF (Reinforcement Learning from Human Feedback), which aligns the raw outputs of Large Language Models with human values and enterprise safety constraints.
See also Large Language Model, Fine-Tuning and Model Inference.