I’m trying to implement a network using Nengo which has synaptic reversal potentials,
Isyn = Wf(Vpre)(Esyn - Vpost),
where Isyn is the synaptic current induced in the postsynaptic neuron, W is the weight, f() is an activation function in the range [0,1], Esyn is the synaptic reversal potential, Vpre and Vpost are the pre and post-synaptic neural voltages.
Is there any way to implement this kind of behavior using Nengo? Or is this not something which Nengo is really built for. To my understanding, it doesn’t really fit with the normal synapse model because there ends up being two neural states multiplied together (W*f(Vpre)*Vpost), and it isn’t really a learning rule.