Concept
MLP
MLP stands for multi-layer perceptron, the plain feedforward neural network: a stack of fully connected (dense) layers with a nonlinearity between them. Think of it as a series of matrix multiplies interleaved with activation functions like ReLU, mapping a fixed-size…
The rest of “MLP” 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→