hidden layer

This term is a technical specification used exclusively within the fields of machine learning and computational neuroscience. It describes the internal processing stages of a neural network that are not directly visible to the external environment, contrasting with the input and output layers.

In a practical architectural context, the number of these layers determines the depth of a model. A network with many such layers is referred to as a deep neural network, where each subsequent layer typically extracts higher-level, more abstract features from the data provided by the previous one.