Spatiotemporal correlations in brain activity are functionally important and have been implicated in perception, learning and plasticity, exploratory behavior, and various aspects of cognition. However, how the collective dynamics of highly recurrent and strongly interacting neurons can evolve into a state in which the activity of individual cells is highly irregular yet macroscopically correlated remains an open fundamental question.
In the talk, I will describe our recent findings on how spatiotemporal correlations can emerge in structured balanced networks and demonstrate how this mechanism might be used in the context of the songbird system to autonomously generate irregular motor behaviors. I will introduce our general theory that relates the strength of pairwise correlations in balanced networks to the network architecture and discuss the key result of the theory: the proof of two theorems according to which macroscopically correlated activity can arise only in specific network structures. The architecture in these networks embeds group of neurons connected to other groups in an effective feedforward manner. I will end the talk by giving several examples of network architecture and discuss how correlations emerge in these network examples.