On the effect of stabilizing mean firing rate of a neuron due to STDP

We show by numerical simulations that a neuron with additive Spike-Timing-Dependent Plasticity with restricted symmetric nearest-neighbor spike pairing scheme, receiving Poisson input, establishes mean firing rate that does not depend on input rates, in a sufficiently high range of input rates. The established rate also does not depend on the number of inputs and the existence of inhibitory inputs, and depends only on the neuron and STDP parameters. A possible way to utilize this effect in learning is shown.