We present and study a probabilistic neural automaton in which the fraction of simultaneously-updated neurons is a parameter, rho (0, 1) . For small rho, there is relaxation towards one of the attractors and a great sensibility to external stimuli an
d, for rho >= rho_c, itinerancy among attractors. Tuning rho in this regime, oscillations may abruptly change from regular to chaotic and vice versa, which allows one to control the efficiency of the searching process. We argue on the similarity of the model behavior with recent observations and on the possible role of chaos in neurobiology.
We studied, both analytically and numerically, complex excitable networks, in which connections are time dependent and some of the nodes remain silent at each time step. More specifically, (a) there is a heterogenous distribution of connection weight
s and, depending on the current degree of order, some connections are reinforced/weakened with strength Phi on short-time scales, and (b) only a fraction rho of nodes are simultaneously active. The resulting dynamics has attractors which, for a range of Phi values and rho exceeding a threshold, become unstable, the instability depending critically on the value of rho. We observe that (i) the activity describes a trajectory in which the close neighborhood of some of the attractors is constantly visited, (ii) the number of attractors visited increases with rho, and (iii) the trajectory may change from regular to chaotic and vice versa as rho is, even slightly modified. Furthermore, (iv) time series show a power-law spectra under conditions in which the attractors space is most efficiently explored. We argue on the possible qualitative relevance of this phenomenology to networks in several natural contexts.
