Thermodynamics for a network of neurons: Signatures of criticality

TL;DR

This paper explores the thermodynamic properties of neural activity patterns in a retina network, revealing signs of criticality that may have functional significance, using experimental data and statistical physics models.

Contribution

It demonstrates the presence of criticality in neural networks through thermodynamic analysis of experimental data, linking neural activity to concepts from statistical physics.

Findings

Signs of a thermodynamic limit in neural activity patterns

Neural network activity is poised near a critical point

Correlation levels influence the critical state of the network

Abstract

The activity of a neural network is defined by patterns of spiking and silence from the individual neurons. Because spikes are (relatively) sparse, patterns of activity with increasing numbers of spikes are less probable, but with more spikes the number of possible patterns increases. This tradeoff between probability and numerosity is mathematically equivalent to the relationship between entropy and energy in statistical physics. We construct this relationship for populations of up to N=160 neurons in a small patch of the vertebrate retina, using a combination of direct and model-based analyses of experiments on the response of this network to naturalistic movies. We see signs of a thermodynamic limit, where the entropy per neuron approaches a smooth function of the energy per neuron as N increases. The form of this function corresponds to the distribution of activity being poised near an unusual kind of critical point. Networks with more or less correlation among neurons would not reach this critical state. We suggest further tests of criticality, and give a brief discussion of its functional significance.