Table 1.

Parameters used in simulations.

Symbol	Description	Value
N	Number of total neurons	{40, 80}
Nin	Number of input neurons	{30, 60}
Nout	Number of output neurons	10
Erev	Reversal potential	30 mV
gL	Decay constant of neuron potential	0.1 μS
τg	Decay constant of synaptic conductances	10 ms
Vthr	Threshold potential for spike emission	1 mV
tref	Refractory period	10 ms
νin	Input firing rate	10 Hz
νtarg	Output target firing rate	{2, 20} Hz
τminfacil	Facilitation time constant - minimum value	1 ms
τmaxfacil	Facilitation time constant - maximum value	000 ms
τminrec	Depression time constant - minimum value	100 ms
τmaxrec	Depression time constant - maximum value	900 ms
Umin	Synaptic utilization - minimum value	0.05
Umax	Synaptic utilization - maximum value	0.95
η	Fixed learning rate for U, τrec, τfacil	0.1
A+2	Amplitude of maximum weights change - pair term in Long-Term Potentiation	4.6 × 10−3
A+3	Amplitude of maximum weights change - triplet term in Long-Term Potentiation	9.1 × 10−3
A−2	Amplitude of maximum weights change - pair term in Long-Term Depression	3.0 × 10−3
A−3	Amplitude of maximum weights change - triplet term in Long-Term Depression	7.5 × 10−9
τm1	Decay constant of presynaptic indicator m1	16.8 ms
τm2	Decay constant of presynaptic indicator m2	575 ms
τo1	Decay constant of postsynaptic indicator o1	33.7 ms
τo2	Decay constant of postsynaptic indicator o2	47 ms
Amin	Lower bound for maximum synaptic weights	0.001
Amax	Higher bound for maximum synaptic weights	1
γ	Learning rate for STDP and for STP-dependent Δw	{1, 2}
dt	Discretization time step	1 ms

STDP parameters are as in the nearest-spike triplet-model, described in Pfister and Gerstner (2006).