Concept
Post-Norm
Post-Norm refers to placing the normalization layer after the residual addition in a transformer block, rather than before it. The block computes x = Norm(x + Sublayer(x)), where Sublayer is attention or the MLP.
The rest of “Post-Norm” 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→