1307.7658 (Daniel B. Larremore et al.)

Inhibition guarantees ceaseless cortex network dynamics    [PDF]

Daniel B. Larremore, Woodrow L. Shew, Shan Yu, Dietmar Plenz, Edward Ott, Francesco Sorrentino, Juan G. Restrepo

Proper function of diverse biological, social, and engineered networks requires a balance of competing influences. In the cerebral cortex, inhibitory neurons keep excitatory neurons in check and play a crucial role in sensory processing. Experiments suggest inhibition is required to maintain cortex network dynamics at the critical point of a phase transition at which neuronal avalanches occur, and further suggest criticality optimizes aspects of cortical information processing. We numerically and analytically investigate critical dynamics in a network model and discover a previously unappreciated consequence of inhibition: the greater the number of inhibitory nodes, the more likely network dynamics are to be intrinsically self-sustaining. In the cortex, ongoing intrinsic dynamics play a role in memory consolidation and learning. We identify critical avalanches and also predict an experimentally observable signature of criticality that we confirm with recordings from the cortex of two awake monkeys, suggesting inhibition confers both balance and ceaseless critical dynamics in the cortex.

View original: http://arxiv.org/abs/1307.7658