forward pass

Feeding an input through every layer of a network in order to produce its final output.

A forward pass computes a network's output by evaluating each layer's neurons in order -- input layer first, then each hidden layer, then the output layer -- feeding each layer's own outputs in as the next layer's inputs, with no step run out of order and no step skipped.

Examples

A two-layer network classifies an image by first computing every neuron in layer 1 from the raw pixels, then feeding those outputs into layer 2 to get the final class scores.