Скачать или смотреть Spiking Neural Network Visualization

This is a demonstration video of my snn library.
So far it is of pretty basic design, the governing rules are as follows:
1. Neurons are represented by black circles
2. Synapses are represented by gray lines
3.When a neuron fires it's action potential runs along connected synapses "axon terminals"
4. A synapse is either inhibitory(black circle) or excitatory(yellow circle), but never both
5. Each neuron's resting potential is -70
6. Each tick every neuron sums it's dendritic connections, if that sum passes it's membrane threshold(-25 to -55 respectively) it fires it's action potential to all axon terminals
7. If a firing neuron causes a post-synaptic neuron to fire, that synapse strengthens, and vice versa
8. Each synapse has a delay modifier(0 to 20), which determines how many ticks will elapse before the post-synaptic neuron will interpret that action potential
9. After a neuron fires it enters a "refractory period" where it rests at -90 and accumulated charge is delayed
10. A synapse may switch form inhibitory to excitatory, or vice versa, if it's strength crosses the origin
11. A tick, the primary measure of time in this simulation, is 100 milliseconds