FLIF

There are a lot of different neural models.
They can be broadly broken into continuous value, spiking point, and spiking compartmental models.
Spiking models emit spikes for communication, continous models send out a value.
Compartmental models break a neuron into compartments and are generally more accurate than point models.
A point model represents a neuron as a point.
FLIF stands for Fatiguing Leaky Integrate and Fire, and it's a spiking point model.
Integrate and Fire models integrate activity from adjacent neurons, and emit a spike if the activity is over a threshold.
Leaky Integrate and Fire models are really common. If the activity doesn't surpass the threshold some of it leaks away; this makes it easier to fire later, but not too easy.
We, rather atypically use a Fatiguing model. If the neuron fires it becomes harder to fire.
The CABots used a fatigue model that only went to 0.
We've more recently allowed it to go below 0, so that a neuron can fire just by being hypo-fatigued.