AI glossary
What is RLHF?
Short answer
Reinforcement learning from human feedback. Uses human preference data to fine-tune model behaviour.
For example
People compare two chatbot answers and pick the more helpful one. A reward model learns those preferences, and the chatbot is then trained to score well on it.
Learn it in a free lesson
- How LLMs are trainedAdvanced · 7 min
- Keeping AI on our side: alignment and interpretabilityAdvanced · 8 min
- Reinforcement learning, deeperAdvanced · 7 min
Related terms
Questions
What is RLHF in simple terms?
Reinforcement learning from human feedback. Uses human preference data to fine-tune model behaviour.
What is an example of RLHF?
People compare two chatbot answers and pick the more helpful one. A reward model learns those preferences, and the chatbot is then trained to score well on it.
Where can I learn about RLHF for free?
AI in Bed explains it in the free lesson "How LLMs are trained" (Advanced level, 7 minutes), with no sign-up.