Growing Critical: Self-Organized Criticality in a Developing Neural System

Kossio, Yaroslav Felipe Kalle; Goedeke, Sven; Akker, Benjamin van den; Ibarz, Borja; Memmesheimer, Raoul-Martin

doi:10.1103/physrevlett.121.058301

Cited by 45 publications

(28 citation statements)

References 70 publications

(111 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We conjecture that the key difference is that doubly stochastic processes propagate the underlying firing rate instead of the actual spike count. Thus, propagation of the actual number of spikes (as e.g., in branching or Hawkes processes; Kossio et al 2018), not some underlying firing rate, seems to be integral to capture the statistics of cortical spiking dynamics.…”

Section: Discussionmentioning

confidence: 99%

Between Perfectly Critical and Fully Irregular: A Reverberating Model Captures and Predicts Cortical Spike Propagation

Wilting

Priesemann

2019

Cerebral Cortex

View full text Add to dashboard Cite

Knowledge about the collective dynamics of cortical spiking is very informative about the underlying coding principles. However, even most basic properties are not known with certainty, because their assessment is hampered by spatial subsampling, i.e., the limitation that only a tiny fraction of all neurons can be recorded simultaneously with millisecond precision. Building on a novel, subsampling-invariant estimator, we fit and carefully validate a minimal model for cortical spike propagation. The model interpolates between two prominent states: asynchronous and critical. We find neither of them in cortical spike recordings across various species, but instead identify a narrow “reverberating” regime. This approach enables us to predict yet unknown properties from very short recordings and for every circuit individually, including responses to minimal perturbations, intrinsic network timescales, and the strength of external input compared to recurrent activation “thereby informing about the underlying coding principles for each circuit, area, state and task.

show abstract

Section: Discussionmentioning

confidence: 99%

Between Perfectly Critical and Fully Irregular: A Reverberating Model Captures and Predicts Cortical Spike Propagation

Wilting

Priesemann

2019

Cerebral Cortex

View full text Add to dashboard Cite

show abstract

“…This raises an interesting question: how does the brain finally reset to a near-critical state after learning, so that another (new) memory can be consolidated? Our work here does not address this issue, but previous work by others has shown that neurons and networks in the brain have built-in homeostatic mechanisms which serves to recalibrate synaptic efficacies (see References [ 48 , 49 , 50 , 51 , 52 ]), a process that was proposed to happen also during development [ 53 ]. Thus, it could be that homeostatic plasticity together with reduced external input during sleep is sufficient to drive the system towards criticality, as shown by Zierenberg, J. et al [ 54 ].…”

Section: Discussionmentioning

confidence: 99%

Critical Dynamics Mediate Learning of New Distributed Memory Representations in Neuronal Networks

Skilling

Ognjanovski

Aton

et al. 2019

Entropy

View full text Add to dashboard Cite

AUTHOR CONTRIBUTION: Q.S., D.M., and M.Z. designed and carried out the computational modeling experiments. N.O. and S.J.A. designed and carried out the behavioral and electrophysiological experiments. N.O. and S.J.A. carried out behavioral analyses and spike sorting to discriminate single-unit activity from electrophysiological recordings. D.M. carried out the analysis of network stability on electrophysiological recordings. Q.S. and M.Z. wrote the paper with assistance from N.O. and S.J.A. Abstract:Critical-state dynamics in the brain have been shown to be an important feature for neural computation. However, links between criticality and network-level mechanisms underlying the formation of new memories are lacking. Here, we record from mice experiencing contextual fear conditioning and probe dynamics for signatures of criticality, finding criticality to be a natural state of the system. We subsequently measure the functional network stability (FuNS) of the recorded neurons and find that tracking changes in the network state before and after fear conditioning accurately predicts fear learning. We turn to modeling to determine the link between critical dynamics and FuNS. We control proximity to a balance between excitation and inhibition, a state we show to be synonymous with criticality, and observe a discrete set of local synaptic changes giving rise to global FuNS only near a critical state. Finally, new information has maximal consolidation potential only near criticality. Author Summary:Much evidence points to the existence and corresponding implications of self-organized critical states in neuronal systems. Here, we expand on this work and show the importance of criticality on the network sensitivity to input and subsequent consolidation of new memories. Our in vivo studies suggest that critical states provide a necessary substrate for network-wide stabilization of functional interactions between neurons. Through modeling, we provide evidence that this socalled functional network stability is most sensitive only near a critical dynamical state. Further, we show that input has global effects on network dynamics only near criticality and, indeed, that new memories can be formed only at these states. Taken together, these results indicate criticalstate dynamics are vital for network sensitivity and consolidation potential.Contextual fear conditioning (CFC) is an optimal experimental paradigm to tackle these questions as it allows for rapid formation and consolidation of memory (i.e. after single-trial learning) 10 . Here, we first characterize hippocampal dynamics in mice subjected to CFC and show that: 1) the hippocampus operates in a near critical regime pre-and post-CFC traininga universal dynamical state indicating a dynamic phase transition, and 2) successful consolidation of fear memory leads to stabilization of network-wide functional representations.While the idea that the brain operates at or near dynamical critically is not new (see for example [11][12][13] and references therein) the functional bene...

show abstract

“…Therefore, due to µ → 1, DMMs must showcase a critical branching process. In fact, in the limit of µ → 1, the Stirling approximation of the Borel distribution is proportional to S −3/2 , namely a scale-free distribution [19,20].…”

Section: Mean-field Analysismentioning

confidence: 99%

Critical branching processes in digital memcomputing machines

Bearden¹,

Sheldon²,

Ventra³

2019

EPL

View full text Add to dashboard Cite

Memcomputing is a novel computing paradigm that employs time non-locality (memory) to solve combinatorial optimization problems. It can be realized in practice by means of non-linear dynamical systems whose point attractors represent the solutions of the original problem. It has been previously shown that during the solution search digital memcomputing machines go through a transient phase of avalanches (instantons) that promote dynamical long-range order. By employing mean-field arguments we predict that the distribution of the avalanche sizes follows a Borel distribution typical of critical branching processes with exponent τ = 3/2. We corroborate this analysis by solving various random 3-SAT instances of the Boolean satisfiability problem. The numerical results indicate a power-law distribution with exponent τ = 1.51 ± 0.02, in very good agreement with the meanfield analysis. This indicates that memcomputing machines self-tune to a critical state in which avalanches are characterized by a branching process, and that this state persists across the majority of their evolution.

show abstract

Growing Critical: Self-Organized Criticality in a Developing Neural System

Cited by 45 publications

References 70 publications

Between Perfectly Critical and Fully Irregular: A Reverberating Model Captures and Predicts Cortical Spike Propagation

Between Perfectly Critical and Fully Irregular: A Reverberating Model Captures and Predicts Cortical Spike Propagation

Critical Dynamics Mediate Learning of New Distributed Memory Representations in Neuronal Networks

Critical branching processes in digital memcomputing machines

Contact Info

Product

Resources

About