ConceptAI & ML
Grokking
A phenomenon where a model suddenly generalizes to unseen examples after running for a long time past the point where training loss has already plateaued.
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
5The lectures that define it or assume it, and the concepts next to it.