Critical Avalanches and Subsampling in Map-based Neural Networks

TL;DR

This paper explores how synaptic noise can induce critical avalanche behavior in neural networks modeled by dynamical maps, highlighting the natural emergence of neuronal properties and the impact of subsampling on detecting criticality.

Contribution

It introduces a novel model using dynamical maps with synaptic noise to study critical avalanches and discusses subsampling effects relevant for experimental detection of criticality.

Findings

Networks of excitatory rebound neurons exhibit power law avalanches.

Dynamical maps naturally produce neuronal properties without imposed assumptions.

Subsampling influences the measurement and detection of critical avalanches.

Abstract

We investigate the synaptic noise as a novel mechanism for creating critical avalanches in the activity of neural networks. We model neurons and chemical synapses by dynamical maps with a uniform noise term in the synaptic coupling. An advantage of utilizing maps is that the dynamical properties (action potential profile, excitability properties, post synaptic potential summation etc.) are not imposed to the system, but occur naturally by solving the system equations. We discuss the relevant neuronal and synaptic properties to achieve the critical state. We verify that networks of excitatory by rebound neurons with fast synapses present power law avalanches. We also discuss the measuring of neuronal avalanches by subsampling our data, shedding light on the experimental search for Self-Organized Criticality in neural networks.