`

Integrate and and Fire

The (leaky) integrate and fire model is perhaps the most prevalent biologically realistic model neuron currently in use. It simulates all the basic aspects of neural conduction in a relatively simple framework.

Simbrain's Integrate and Fire neuron is a spiking neuron. As a result, when it spikes, a ring of yellow flashes around it and the axon protruding from it turns yellow as well.

Activation of the neuron is interpreted as voltage potential. When that voltage potential passes a threshold value, the neuron spikes (resulting in the flash of yellow), and then the voltage is reset to a reset value, typically below the threshold potential. Below threshold, the neuron behaves like the additive neuron. It decays to a steady state value at a rate proporitional to its time constant.

Activation is computed by integrating the following differential equation using Euler's method:

 

where r is resting potential, R is resistance, τ is a time constant, and W is the weighted inputs.

 

Time step

See time-step.

Resting potential

The activation value which the neuron decays to over time.

Reset potential

The activation value which the neuron is reset to after a spiking event.

Threshold

The value beyond which the neuron spikes.

Time constant

Controls the rate at which the neuron attains its resting potential. As one increases this value, things become relatively slower; conversely, as one decreases this value, things become relatively faster.

Resistance

A constant factor multiplied by weighted inputs.

Use Clipping

See clipping.

Add Noise

If this is set to true, random values are added to the activation via a noise generator.