SELF-SUSTAINED FIRING IN POPULATIONS OF INTEGRATE-AND-FIRE NEURONS

A coupled population of simple integrate-and-fire model neurons is analyzed to determine the conditions that lead to stable firing sustained by all-to-all excitatory interactions. The possible temporal firing patterns are determined, and the initial conditions that produce them are discussed. For certain parameter values, two different patterns of firing are possible: one in which the firing is distributed among many groups of neurons that fire sequentially, and another in which the population breaks into two groups that fire alternately. The probabilities of different temporal firing patterns arising from random initial conditions and noise inputs are analyzed by computer simulation.