Figure 6.

Effects of adaptation on resonance properties of recurrent networks. Gain (top) and phase shift (bottom) of the spike rate r_ℰ for networks with dominant recurrent excitation [g = 1.05, (A)] and inhibition [g = 1.5, (B)] as a function of the input frequency f. The gain is defined as the quotient of the oscillation amplitude in r_ℰ for the input with frequency f and the amplitude for the lowest frequency (f_min = 0.5 Hz). Adaptation parameter values are a = b = 0 (black), a = 5 nS, b = 0.05 nA, τ_w = 100 ms (dark red), and a = 5 nS, b = 0.05 nA, τ_w = 400 ms (orange). Delay distributions are identically parameterized (p^ℰ_d ≡ p^ℐ_d) with τ_r = 1.25 ms and τ_d = 1.5 ms. The Poisson-input rates r^ℰ_ext, r^ℐ_ext each consist of a baseline rate plus a sinusoidal component of small amplitude (1/1000th of the baseline) with frequency f. The baseline of r^ℰ_ext is chosen to yield a steady-state spike rate r^∞_ℰ of 50 Hz with constant input rate. The baseline rate of r^ℐ_ext is chosen as explained in the Methods section.