Dynamical Entropy Production in Spiking Neuron Networks in the Balanced State

TL;DR

This paper demonstrates that large sparse networks of theta neurons in the balanced state exhibit deterministic extensive chaos, with high entropy production limiting cortical spike pattern information capacity.

Contribution

It provides the first numerical calculation of Lyapunov spectra, entropy production, and attractor dimension in such neural networks, revealing the chaotic dynamics.

Findings

Extensive chaos observed in inhibitory networks.

Chaos intensity increases with excitatory populations.

High entropy production constrains information encoding.

Abstract

We demonstrate deterministic extensive chaos in the dynamics of large sparse networks of theta neurons in the balanced state. The analysis is based on numerically exact calculations of the full spectrum of Lyapunov exponents, the entropy production rate and the attractor dimension. Extensive chaos is found in inhibitory networks and becomes more intense when an excitatory population is included. We find a strikingly high rate of entropy production that would limit information representation in cortical spike patterns to the immediate stimulus response.