ConceptAI & ML
Model-based RL with human preferences
An RL approach where a reward model is trained less frequently on preference data, and many RL updates are subsequently made on that same reward model.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Lecturers
1Connections
3The lectures that define it or assume it, and the concepts next to it.