What does it mean that a neuron fires, in layman terms? What is the output usually?

perigk profile image Periklis Gkolias ・1 min read

I am reading some texts for deep learning, but everyone seems to take it as self-explanatory. :)

My understanding is that the activation function gets a WX + b which equals, lets say 1.745, as an input and returns a...value. What does this value represent? Is it just a new "weight"? Is it something else?


markdown guide

A Neuron firing basically implies if it'll be "activated" or not i.e. if it'll be able to output something that affects further calculation. Consider the example you gave above, if WX + b is 1.745, this value is passed into an activation function which finally decides the output value of this neuron. In a simple case, assume that this activation function is a simple binary step function then the output value becomes 1.
Hope this helps :)


"Neurons that fire together wire together"?

From TensorFlow homepage: "Next, the network is asked to solve a problem, which it attempts to do over and over, each time strengthening the connections that lead to success and diminishing those that lead to failure."

Are you looking for more technical details or an abstract concept?


After reading more materials I realized that it is just an indicator of the input(eg maybe a probability) and the next layer knows how to deal with it.

Thanks :)


In neuroscience terms it means that the the threshold for an action potential (name of a process) has been passed (40mv) then the neuron will fire and an electrical impulse will propagate.