Tailored ensembles of neural networks optimize sensitivity to stimulus statistics

Zierenberg, Johannes; Wilting, Jens; Priesemann, Viola; Levina, Anna

doi:10.1103/physrevresearch.2.013115

Cited by 28 publications

(23 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One way in which the brain can take advantage of the information processing benefits of different states is to regulate its dynamics with respect to the critical point in a task-dependent manner (Pfeffer et al, 2018). Zierenberg et al (2020) show that the brain could also combine networks with different distances to the critical point, such that the resulting ensemble has a dynamic range of stimulus response wider than any of the ensemble networks. This strategy could plausibly be employed to extend the range of stimuli strengths over which pre-stimulus amplitude and phase regulation occur.…”

Section: Discussionmentioning

confidence: 99%

Pre-stimulus phase and amplitude regulation of phase-locked responses are maximized in the critical state

Avramiea

Hardstone

Lueckmann

et al. 2020

eLife

View full text Add to dashboard Cite

Understanding why identical stimuli give differing neuronal responses and percepts is a central challenge in research on attention and consciousness. Ongoing oscillations reflect functional states that bias processing of incoming signals through amplitude and phase. It is not known, however, whether the effect of phase or amplitude on stimulus processing depends on the long-term global dynamics of the networks generating the oscillations. Here, we show, using a computational model, that the ability of networks to regulate stimulus response based on pre-stimulus activity requires near-critical dynamics—a dynamical state that emerges from networks with balanced excitation and inhibition, and that is characterized by scale-free fluctuations. We also find that networks exhibiting critical oscillations produce differing responses to the largest range of stimulus intensities. Thus, the brain may bring its dynamics close to the critical state whenever such network versatility is required.

show abstract

Section: Discussionmentioning

confidence: 99%

Pre-stimulus phase and amplitude regulation of phase-locked responses are maximized in the critical state

Avramiea

Hardstone

Lueckmann

et al. 2020

eLife

View full text Add to dashboard Cite

show abstract

“…[7,[21][22][23][24][25][26][27]). One of the main results is that information processing seems to be optimized at a secondorder absorbing phase transition [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42]. This transition occurs between no activity (the absorbing phase) and nonzero steadystate activity (the active phase).…”

Section: Introductionmentioning

confidence: 99%

Mechanisms of Self-Organized Quasicriticality in Neuronal Network Models

2020

View full text Add to dashboard Cite

The critical brain hypothesis states that there are information processing advantages for neuronal networks working close to the critical region of a phase transition. If this is true, we must ask how the networks achieve and maintain this critical state. Here, we review several proposed biological mechanisms that turn the critical region into an attractor of a dynamics in network parameters like synapses, neuronal gains, and firing thresholds. Since neuronal networks (biological and models) are not conservative but dissipative, we expect not exact criticality but self-organized quasicriticality, where the system hovers around the critical point.

show abstract

“…Using the branching process as a simple model of neuronal activity, it is intuitive to think of the intrinsic timescale as the duration over which any perturbation reverberates (or persists) within the network [ 13 , 45 ]. According to this intuition, different timescales should benefit different functional aspects of cortical networks [ 12 , 46 , 47 ].…”

Section: Discussionmentioning

confidence: 99%

MR. Estimator, a toolbox to determine intrinsic timescales from subsampled spiking activity

et al. 2021

Self Cite

View full text Add to dashboard Cite

Here we present our Python toolbox “MR. Estimator” to reliably estimate the intrinsic timescale from electrophysiologal recordings of heavily subsampled systems. Originally intended for the analysis of time series from neuronal spiking activity, our toolbox is applicable to a wide range of systems where subsampling—the difficulty to observe the whole system in full detail—limits our capability to record. Applications range from epidemic spreading to any system that can be represented by an autoregressive process. In the context of neuroscience, the intrinsic timescale can be thought of as the duration over which any perturbation reverberates within the network; it has been used as a key observable to investigate a functional hierarchy across the primate cortex and serves as a measure of working memory. It is also a proxy for the distance to criticality and quantifies a system’s dynamic working point.

show abstract

Tailored ensembles of neural networks optimize sensitivity to stimulus statistics

Cited by 28 publications

References 48 publications

Pre-stimulus phase and amplitude regulation of phase-locked responses are maximized in the critical state

Pre-stimulus phase and amplitude regulation of phase-locked responses are maximized in the critical state

Mechanisms of Self-Organized Quasicriticality in Neuronal Network Models

MR. Estimator, a toolbox to determine intrinsic timescales from subsampled spiking activity

Contact Info

Product

Resources

About