Weak pairwise correlations imply strongly correlated network states in a neural population

@article{Schneidman2006WeakPC,
  title={Weak pairwise correlations imply strongly correlated network states in a neural population},
  author={Elad Schneidman and Michael J. Berry and Ronen Segev and William Bialek},
  journal={Nature},
  year={2006},
  volume={440},
  pages={1007-1012}
}
Biological networks have so many possible states that exhaustive sampling is impossible. Successful analysis thus depends on simplifying hypotheses, but experiments on many systems hint that complicated, higher-order interactions among large groups of elements have an important role. Here we show, in the vertebrate retina, that weak correlations between pairs of neurons coexist with strongly collective behaviour in the responses of ten or more neurons. We find that this collective behaviour is… 

Figures and Topics from this paper

Spin glass models for a network of real neurons
TLDR
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.
When are correlations strong
TLDR
An expansion for the entropy of Ising systems in powers of the correlations is developed, carrying this out to fourth order, and it is shown that even though all pairwise correlations are weak, the fact that these correlations are widespread means that their impact on the network as a whole is not captured in the leading orders of perturbation theory.
When are microcircuits well-modeled by maximum entropy methods?
TLDR
Overall, the findings identify circuit-level mechanisms that produce and fail to produce higher-order spiking statistics in neural ensembles.
Correlations and Population Dynamics in Cortical Networks
TLDR
It is shown that simple mean-field models that take the structure of the coupling matrix into account can adequately describe the power spectra of the population activity and the consequences of Dale's principle on correlations and rate fluctuations are discussed in the light of recent experimental findings.
Sparse low-order interaction network underlies a highly correlated and learnable neural population code
TLDR
It is shown that because of the sparse nature of the neural code, the higher-order interactions can be easily learned using a novel model and that a very sparse low-order interaction network underlies the code of large populations of neurons.
A second-order maximum entropy model predicts correlated network states, but not their evolution over time
TLDR
It is concluded that a second- order maximum entropy model can predict correlated states, but not their evolution over time, which suggests that higher-order maximum entropy models incorporating temporal interactions will be needed to account for the sequences of correlated states that are often observed in the data.
High-Order Correlations Explain the Collective Behavior of Cortical Populations in Executive, But Not Sensory Areas
TLDR
Surprisingly, despite the fact the firing pattern of neurons is controlled by brain state, the model based on pairwise interactions captures more than 90% of the structure in the detailed patterns of spiking observed during wakefulness and sleep.
Searching for Collective Behavior in a Large Network of Sensory Neurons
TLDR
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.
How fast can we learn maximum entropy models of neural populations
Most of our knowledge about how the brain encodes information comes from recordings of single neurons. However, computations in the brain are carried out by large groups of neurons. Modelling the
A Simple Mechanism for Beyond-Pairwise Correlations in Integrate-and-Fire Neurons
TLDR
This work demonstrates the strong connection between the collective dynamics produced by integrate-and-fire and dichotomous Gaussian models, and shows that the latter is a surprisingly accurate model of the former.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 60 REFERENCES
Nonlinear Population Codes
TLDR
This letter studies correlated neuronal populations whose higher-order statistics, and in particular response variances, are also modulated by the stimulus, and proposes a bilinear readout scheme that involves spatial decorrelation, quadratic nonlinearity, and population vector summation.
Correlated neuronal discharge rate and its implications for psychophysical performance
TLDR
To assess correlation within a pool, responses of pairs of neurons were recorded simultaneously during repeated stimulus presentations and it is reported that the observed covariation in spike count was relatively weak, the correlation coefficient averaging 0.12.
Neural Coding: Higher-Order Temporal Patterns in the Neurostatistics of Cell Assemblies
TLDR
This work presents test statistics for detecting the presence of higher-order interactions in spike train data by parameterizing these interactions in terms of coefficients of log-linear models and presents a Bayesian approach for inferring the existence or absence of interactions and estimating their strength.
Dynamics of neuronal interactions in monkey cortex in relation to behavioural events
TLDR
It is shown that correlated firing between single neurons, recorded simultaneously in the frontal cortex of monkeys performing a behavioural task, evolves within a fraction of a second, and in systematic relation to behavioural events, support the notion that neurons can associate rapidly into a functional group in order to perform a computational task, at the same time becoming dissociated from concurrently activated competing groups.
Redundancy in the Population Code of the Retina
TLDR
A high degree of retinal redundancy suggests that design principles beyond coding efficiency may be important at the population level.
Neural networks and physical systems with emergent collective computational abilities.
  • J. Hopfield
  • Computer Science, Medicine
    Proceedings of the National Academy of Sciences of the United States of America
  • 1982
TLDR
A model of a system having a large number of simple equivalent components, based on aspects of neurobiology but readily adapted to integrated circuits, produces a content-addressable memory which correctly yields an entire memory from any subpart of sufficient size.
Multineuronal Firing Patterns in the Signal from Eye to Brain
Population codes in the brain have generally been characterized by recording responses from one neuron at a time. This approach will miss codes that rely on concerted patterns of action potentials
Synaptic modification by correlated activity: Hebb's postulate revisited.
  • G. Bi, M. Poo
  • Psychology, Medicine
    Annual review of neuroscience
  • 2001
TLDR
Spike timing-dependent modifications, together with selective spread of synaptic changes, provide a set of cellular mechanisms that are likely to be important for the development and functioning of neural networks.
Selective Adaptation in Networks of Cortical Neurons
TLDR
The observation that selective adaptation arises naturally in a network of cortical neurons developing ex vivo indicates that this is an inherent feature of spontaneously organizing cortical networks.
Dynamics of neuronal firing correlation: modulation of "effective connectivity".
TLDR
Adopting a model-based approach, this work develops procedures to quantify and properly normalize the classical joint peristimulus time scatter diagram and generalizes the classical measures for quantifying a direct interneuronal connection to include possible stimulus-locked time variations.
...
1
2
3
4
5
...