Concept
Advantage
Advantage is the quantity in reinforcement learning that answers "how much better than expected was this action?" Formally it is A = Q(s,a) - V(s), the action-value minus the state-value baseline. You already know the intuition from variance reduction: raw returns tell you…
The rest of “Advantage” is a premium feature: every concept in the library gets a precise, practitioner-focused write-up like this one, cross-linked straight from the paper summaries that use it.
Log in to unlock→