Linear response for spiking neuronal networks with unbounded memory

摘要: We establish a general linear response relation for spiking neuronal networks, based on chains with unbounded memory. This allows us to predict the influence of weak amplitude time-dependent external stimuli spatio-temporal spike correlations, from spontaneous statistics (without stimulus) in context where memory dynamics can extend arbitrarily far past. Using this approach, we show how is explicitly related an example, gIF model, introduced by M. Rudolph and A. Destexhe. example illustrates collective effect stimuli, intrinsic dynamics, network connectivity statistics. illustrate our results numerical simulations.