Echoes in correlated neural systems

TL;DR

This paper develops a quantitative theory for pairwise correlations in finite-sized random networks of spiking neurons, explaining complex neural dynamics and collective oscillations with explicit analytic expressions.

Contribution

It introduces a novel analytical framework for understanding correlations in directed, time-delayed neural networks, surpassing mean field limitations.

Findings

Derived explicit formulas for cross correlation functions.

Explained the impact of single action potentials on neural dynamics.

Provided criteria for the emergence of collective oscillations.

Abstract

Correlations are employed in modern physics to explain microscopic and macroscopic phenomena, like the fractional quantum Hall effect and the Mott insulator state in high temperature superconductors and ultracold atoms. Simultaneously probed neurons in the intact brain reveal correlations between their activity, an important measure to study information processing in the brain that also influences macroscopic signals of neural activity, like the electro encephalogram (EEG). Networks of spiking neurons differ from most physical systems: The interaction between elements is directed, time delayed, mediated by short pulses, and each neuron receives events from thousands of neurons. Even the stationary state of the network cannot be described by equilibrium statistical mechanics. Here we develop a quantitative theory of pairwise correlations in finite sized random networks of spiking neurons. We derive explicit analytic expressions for the population averaged cross correlation functions. Our theory explains why the intuitive mean field description fails, how the echo of single action potentials causes an apparent lag of inhibition with respect to excitation, and how the size of the network can be scaled while maintaining its dynamical state. Finally, we derive a new criterion for the emergence of collective oscillations from the spectrum of the time-evolution propagator.