Machine Learning
Glossary
Reinforcement Learning
Training an agent to take actions that maximise a reward through trial and error.
Last updated
In reinforcement learning, an agent interacts with an environment, receives rewards and adjusts its strategy to earn more over time. It has driven breakthroughs in games and robotics. The challenge is designing a reward that encourages the behaviour you actually want, since agents become very good at exploiting loopholes.
Related terms
RLHF
Reinforcement Learning from Human Feedback, a method that aligns models with human preferences using ranked examples.
AI Agent
A system that uses a model to plan and carry out multi-step tasks, often calling tools along the way.
Machine Learning
A branch of AI in which models learn patterns from data instead of being explicitly programmed with rules.
Keep exploring
Browse the full AI glossary or compare AI tools that use this technology.