Core Concepts

What Is RLHF (Reinforcement Learning from Human Feedback)?

Reinforcement learning from human feedback (RLHF) is a training method that uses human preferences to fine-tune a model’s behavior. People rank model outputs, a reward model learns those preferences, and the model is optimized against it. RLHF was central to making ChatGPT and similar assistants follow instructions and behave more helpfully.