Spiking Network Model Builder

To build a biologically realistic spiking neural network, you specify neuron models, synapse dynamics, connectivity patterns, E/I balance, and plasticity rules. This skill guides the complete parameter selection process and validates output statistics like firing rates and synchrony.

STDP modifies synaptic weights based on relative timing of pre- and post-synaptic spikes in spiking networks. This skill configures STDP, BCM, and homeostatic plasticity rules alongside neuron models and synapse dynamics to simulate adaptive cortical microcircuits.

Choosing a neuron model depends on biological accuracy and computational constraints for your spiking network. This skill guides selection between LIF, EIF, AdEx, Izhikevich, and HH models, detailing when to use each for cortical microcircuits or AI-regime networks.

Yes, this spiking network builder works with NEST, Brian2, NEURON, and GeNN. It provides recommended toolchains and validates simulation parameters against cortical statistics across these major neural simulation platforms.

Configuring E/I balance and weight scaling in a spiking neural network requires setting explicit excitatory and inhibitory neuron ratios, such as an 80/20 split, and adjusting synaptic weights. This skill enforces these parameters to create stable balanced AI networks and prevent runaway firing.