Concept
Top-k gating
Top-k gating is a routing mechanism that sends each input to only the k highest-scoring experts out of many, rather than running all of them. Think of it like a softmax classifier, except instead of producing a full probability distribution you keep only the top k…
The rest of “Top-k gating” 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→