• Corpus ID: 9208533

Thermodynamics for a network of neurons: Signatures of criticality

  title={Thermodynamics for a network of neurons: Signatures of criticality},
  author={Ga{\vs}per Tka{\vc}ik and Thierry Mora and Olivier Marre and Dario Amodei and Michael J. Berry and William Bialek},
  journal={arXiv: Neurons and Cognition},
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… 

Figures from this paper

Dynamical criticality in the collective activity of a population of retinal neurons.
A novel approach to assess criticality that overcomes limitations of arbitrary order parameters, while encompassing and generalizing previous criteria, and takes into account the temporal dynamics of the activity greatly enhances the evidence for criticality.
On the ground state of spiking network activity in mammalian cortex
Deploying a novel, subsampling-invariant estimator, it is found that in vivo dynamics do not comply with either hypothesis, but instead occupy a narrow "reverberating" state consistently across multiple mammalian species and cortical areas.
Energetics of stochastic BCM type synaptic plasticity and storing of accurate information
It is shown that synaptic weights in an extended version of BCM plasticity with a decay term, stochastic noise, and nonlinear dependence of neuron’s firing rate on synaptic current exhibit bistability, which suggests that synaptic information is not directly limited by energy consumption.
Testing the critical brain hypothesis using a phenomelogical renormalization group
We present a systematic study to test a recently introduced phenomenological renormalization group, proposed to coarse-grain data of neural activity from their correlation matrix. The approach
A Moment-Based Maximum Entropy Model for Fitting Higher-Order Interactions in Neural Data
The Reliable Moment model is extended on the Reliable Interaction class of models to develop a normalized variant that adaptively identifies the specific pairwise and higher-order moments that can be estimated from a given dataset for a specified confidence level and is able to capture cortical-like distributions of population spiking patterns.
Thermodynamic Formalism in Neuronal Dynamics and Spike Train Statistics
How the Thermodynamic Formalism can be exploited in the field of theoretical neuroscience, as a conceptual and operational tool, in order to link the dynamics of interacting neurons and the statistics of action potentials from either experimental data or mathematical models is reviewed.
Large Deviations Properties of Maximum Entropy Markov Chains from Spike Trains
This work provides an accessible introduction to the maximum entropy Markov chain inference problem and large deviations theory to the community of computational neuroscience, avoiding some technicalities while preserving the core ideas and intuitions.
An Introduction to the Non-Equilibrium Steady States of Maximum Entropy Spike Trains
A detailed account of the mathematical foundations and work out examples to illustrate the key concepts and results from non-equilibrium statistical mechanics are provided.
Statistical mechanics of consciousness: Maximization of information content of network is associated with conscious awareness.
A surprisingly simple result is found: Normal wakeful states are characterized by the greatest number of possible configurations of interactions between brain networks, representing highest entropy values, suggesting that consciousness could be the result of an optimization of information processing.
Strong Correlations in gravity and biophysics
The unifying theme of this dissertation is the use of correlations. In the first part (chapter 2), we investigate correlations in quantum field theories in de Sitter space. In the second part


Searching for Collective Behavior in a Large Network of Sensory Neurons
The properties of the neural vocabulary are explored by estimating its entropy, which constrains the population's capacity to represent visual information, and classifying activity patterns into a small set of metastable collective modes, showing that the neural codeword ensembles are extremely inhomogenous.
The simplest maximum entropy model for collective behavior in a neural network
This work analyzes the responses of neurons in a small patch of the retina to naturalistic stimuli, and finds that the implied thermodynamics is very close to an unusual critical point, in which the entropy is exactly equal to the energy.
Spin glass models for a network of real neurons
It is shown that Pairwise interactions between neurons account for observed higher-order correlations, and that for groups of 10 or more neurons pairwise interactions can no longer be regarded as small perturbations in an independent system.
Neuronal Avalanches in Neocortical Circuits
This work shows that propagation of spontaneous activity in cortical networks is described by equations that govern avalanches, and suggests that “neuronal avalanches” may be a generic property of cortical networks, and represent a mode of activity that differs profoundly from oscillatory, synchronized, or wave-like network states.
Statistical mechanics of neural networks near saturation
Stimulus-dependent Maximum Entropy Models of Neural Population Codes
It is shown how the stimulus-dependent maximum entropy model, in conjunction with static maximum entropy models of population vocabulary, can be used to estimate information-theoretic quantities like average surprise and information transmission in a neural population.
Being Critical of Criticality in the Brain
The concept of criticality is explained and substantial objections to the criticality hypothesis raised by skeptics are reviewed, and counter points are presented in dialog form.
Universal critical dynamics in high resolution neuronal avalanche data.
It is shown that the dynamics of cultured cortical networks are critical, and the mean temporal profiles of avalanches of widely varying durations are quantitatively described by a single universal scaling function.
Ising models for networks of real neurons
Ising models with pairwise interactions are the least structured, or maximum-entropy, probability distributions that exactly reproduce measured pairwise correlations between spins, and here they are constructed that describe the correlated spiking activity of populations of 40 neurons in the retina.
The Structure of Multi-Neuron Firing Patterns in Primate Retina
Large-scale multi-electrode recordings were used to measure electrical activity in nearly complete, regularly spaced mosaics of several hundred ON and OFF parasol retinal ganglion cells in macaque monkey retina, and pairwise and adjacent interactions accurately accounted for the structure and prevalence of multi-neuron firing patterns.