Build a Full SOEN Neuron

A build-it-up interactive model of a superconducting optoelectronic network across eight stages. From the SQUID device transfer function you progressively switch on a synapse that turns detected photons into stored flux, a dendrite that leaky-integrates synaptic signal, a soma that fires a spike at threshold and recovers through a refractory loop, and an optical fan-out arbor that broadcasts each spike to many targets. The next stages zoom out to a rate-model network (two inputs, six hidden, one output) trained on XOR by equilibrium propagation, a recurrent Hopfield network that recalls a stored pattern from a corrupted input as its energy descends, and a power-budget comparison of synaptic operations per watt for a cryogenic SOEN system versus a GPU farm at the same wall-plug power.