Figure 6
From: Mesoscopic population equations for spiking neural networks with synaptic short-term plasticity
Recurrent network of Poisson neurons with finite-size noise generating irregular population spikes. (A) For small population size, \(N=100\), microscopic (i) as well as first- (ii) and second-order (iii) MF dynamics exhibit irregular population spikes. For large population size, \(N=5000\), and hence weak finite-size noise, population spikes cease in microscopic (iv) and second-order MF dynamics (vi) consistent with an excitable dynamics, whereas the first-order MF approximation (v) wrongly predicts regular population spikes corresponding to an underlying oscillatory (limit-cycle) dynamics. (B) Time-averaged population activity decreases with increasing population size indicating a decrease of population spike frequency. The prediction of the second-order MF is accurate across all population sizes, which is not the case with the first-order MF, especially for large populations. (C) PSD of population activities for \(N=100\) neurons. Model parameters are detailed in Appendix E. Superscript (N) of \(A^{N}\) is omitted in the legends