Perceptrons and feedback circuits

In standard mathematical logic, a logic gate activates, or 'fires' depending on whether certain inputs are either ON of OFF

Example of a logical AND gate.

In an artificial neural network, this concept is extended: inputs can carry differing levels of current so that a gate (neurone) fires provided that the overall input level exceeds a certain threshold.

For what ranges of input levels will these neurones fire?


SOME EXAMPLES

SOME PROBLEM WITH XOR

A feedback circuit

ACTIVATION FUNCTION -- SIGMOIDS