Concept
reverse KL divergence
Reverse KL divergence measures how far one probability distribution q sits from a reference p, computed as KL(q || p) = E_{x~q}[log q(x) - log p(x)]. It is the "reverse" of the more familiar forward KL, KL(p || q), which is what maximum likelihood training minimizes.
The rest of “reverse KL divergence” 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→