Hidden Layer

Appears in 1 paper

A layer of neurons between the input layer and the output layer.

As used in Paper 03 — Learning Representations by Back-propagating Errors →

A layer of neurons between the input layer and the output layer. "Hidden" because their outputs are not directly observable — they are intermediate representations used internally by the network. The existence of hidden layers is what allows networks to learn complex, non-linear patterns. Credit assignment to hidden layers is what backpropagation solves.