ConceptAI & ML
Eligibility traces in policy gradient
An eligibility trace for policy gradients maintains a memory of recent action scores to allow updates based on multi-step returns, behaving analogously to value-based TD lambda.
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.