None: Boosting of neural circuit chaos at the onset of collective oscillations

Agostina Palmigiano, Rainer Engelken, Fred WolfOpenalex Unknown 2023 · DOI · OpenAlex

Abstract: Neuronal spiking activity in cortical circuits is often temporally structured by collective rhythms. Rhythmic activity has been hypothesized to regulate temporal coding and to mediate the flexible routing of information flow across the cortex. Spiking neuronal circuits, however, are non-linear systems that, through chaotic dynamics, can amplify insignificant microscopic fluctuations into network-scale response variability. In nonlinear systems in general, rhythmic oscillatory drive can induce chaotic behavior or boost the intensity of chaos. Thus, neuronal oscillations could rather disrupt than facilitate cortical coding functions by flooding the finite population bandwidth with chaotically-boosted noise. Here we tackle a fundamental mathematical challenge to characterize the dynamics on the attractor of effectively delayed network models. We find that delays introduce a transition to collective oscillations, below which ergodic theory measures have a stereotypical dependence on the delay so far only described in scalar systems and low-dimensional maps. We demonstrate that the emergence of internally generated oscillations induces a complete dynamical reconfiguration, by increasing the dimensionality of the chaotic attractor, the speed at which nearby trajectories separate from one another, and the rate at which the network produces entropy. We find that periodic input drive leads to a dramatic increase of chaotic measures at a the resonance frequency of the recurrent network. However, transient oscillatory input only has a moderate role on the collective dynamics. Our results suggest that simple temporal dynamics of the mean activity can have a profound effect on the structure of the spiking patterns and therefore on the information processing capability of neuronal networks.

Introduction

Our brains are noisy, chaotic systems, yet they frequently exhibit highly structured collective rhythms. For decades, neuroscientists have debated whether these oscillations help organize neural firing to transmit information more effectively or if they are simply a byproduct of circuit dynamics. A major hurdle in answering this has been mathematical: real neurons have synaptic delays, which turn a network into an "infinite-dimensional" system. This makes it notoriously difficult to calculate the Lyapunov spectrum—the standard toolkit for measuring chaos. Without these tools, it has been impossible to precisely determine whether oscillations quench chaos or intensify it.

Methods

The researchers introduced a novel "two-stage" neuronal model to bypass the infinite-dimension problem. Each neuron consists of a soma (modeled as a quadratic integrate-and-fire unit) and a Single Compartment Axon (SCA). Instead of treating the delay as a history function, the axon is modeled as an excitable unit that relays the spike after a set time. This innovation maps the system back into a finite, $2N$-dimensional space, allowing for the semi-analytical computation of the entire spectrum of Lyapunov exponents. From these, the team derived the attractor dimension (the number of degrees of freedom in the system) and metric entropy (the rate of noise production).

Key Insights and Results

The study reveals that at the onset of collective oscillations (a Hopf bifurcation), measures of chaos—including the maximum Lyapunov exponent, attractor dimension, and metric entropy—all intensify. This "chaos boost" suggests that oscillations can flood the network with variability rather than taming it.

Asynchronous state (delta=0) characterized by flat mean activity and a baseline dimensionality proportional to the number of neurons.

Prior to the onset of oscillations, the dimensionality of the chaotic attractor increases linearly with the synaptic delay. However, the emergence of internally generated oscillations induces a complete dynamical reconfiguration, significantly increasing the speed at which nearby trajectories diverge.

Oscillatory state (delta != 0) showing rhythmic mean activity and the transition to a more complex, higher-dimensional chaotic attractor.

Key findings include:

  • Spiking vs. Rate Models: This boosting is unique to spiking networks. In simpler rate-based models, oscillations often cause chaos to collapse, suggesting that the discrete nature of spikes is fundamental to these dynamics.
  • Heterogeneity as a Safeguard: In uniform networks, long delays can cause a collapse into a perfectly repeating limit cycle. Single-cell heterogeneity prevents this "pathological" synchronization, allowing chaos to persist even during strong oscillations.
  • Resonance and Transients: Chaos is most heavily boosted when an external drive matches the network's natural resonant frequency. However, transient oscillations—common in the real cortex—have only a moderate effect, suggesting the brain may use short bursts to communicate without triggering a massive noise flood.

Discussion

The findings suggest that simple temporal dynamics of mean activity can profoundly affect spiking patterns and information processing. While the two-stage model is a robust proxy, future work must bridge the gap between abstract mathematical measures like the Kaplan-Yorke dimension and experimental observations. The central question remains: does this boosted chaos facilitate complex computations, or is it a hurdle the brain has evolved to manage?

Summary updated

Abstract

Neuronal spiking activity in cortical circuits is often temporally structured by collective rhythms. Rhythmic activity has been hypothesized to regulate temporal coding and to mediate the flexible routing of information flow across the cortex. Spiking neuronal circuits, however, are non-linear systems that, through chaotic dynamics, can amplify insignificant microscopic fluctuations into network-scale response variability. In nonlinear systems in general, rhythmic oscillatory drive can induce chaotic behavior or boost the intensity of chaos. Thus, neuronal oscillations could rather disrupt than facilitate cortical coding functions by flooding the finite population bandwidth with chaotically-boosted noise. Here we tackle a fundamental mathematical challenge to characterize the dynamics on the attractor of effectively delayed network models. We find that delays introduce a transition to collective oscillations, below which ergodic theory measures have a stereotypical dependence on the delay so far only described in scalar systems and low-dimensional maps. We demonstrate that the emergence of internally generated oscillations induces a complete dynamical reconfiguration, by increasing the dimensionality of the chaotic attractor, the speed at which nearby trajectories separate from one another, and the rate at which the network produces entropy. We find that periodic input drive leads to a dramatic increase of chaotic measures at a the resonance frequency of the recurrent network. However, transient oscillatory input only has a moderate role on the collective dynamics. Our results suggest that simple temporal dynamics of the mean activity can have a profound effect on the structure of the spiking patterns and therefore on the information processing capability of neuronal networks.
BibTeX
@inproceedings{palmigiano2023none,
  author = {Agostina Palmigiano and Rainer Engelken and Fred Wolf},
  title = {{None: Boosting of neural circuit chaos at the onset of collective oscillations}},
  booktitle = {Openalex Unknown},
  year = {2023},
  doi = {10.7554/elife.90378.1.sa2},
}

Loading related content...

Figure