Signal Propagation and Logic Gating in Networks of Integrate-and-Fire Neurons

By Tim P. Vogels, L. F. Abbott. Journal of Neuroscience, 2005

509 citations recorded by the source corpus.

Transmission of signals within the brain is essential for cognitive function, but it is not clear how neural circuits support reliable and accurate signal propagation over a sufficiently large dynamic range. Two modes of propagation have been studied: synfire chains, in which synchronous activity travels through feedforward layers of a neuronal network, and the propagation of fluctuations in firing rate across these layers. In both cases, a sufficient amount of noise, which was added to previous models from an external source, had to be included to support stable propagation. Sparse, randomly connected networks of spiking model neurons can generate chaotic patterns of activity. We investigate whether this activity, which is a more realistic noise source, is sufficient to allow for signal transmission. We find that, for rate-coded signals but not for synfire chains, such networks support robust and accurate signal reproduction through up to six layers if appropriate adjustments are made in synaptic strengths. We investigate the factors affecting transmission and show that multiple signals can propagate simultaneously along different pathways. Using this feature, we show how different types of logic gates can arise within the architecture of the random network through the strengthening of specific synapses.

Subfields: Cognitive Neuroscience, Electrical and Electronic Engineering, Statistical and Nonlinear Physics

View Paper: https://www.semanticscholar.org/paper/d90da710a64bc291f59d17f0a0ce58375e6a5d55

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Electrical and Electronic Engineering

In Statistical and Nonlinear Physics

Citations in the corpus

Training Spiking Neural Networks Using Lessons From Deep Learning (2021, background): “The input spike train undergoes two rounds of “smoothing,” which means that the membrane potential has a finite rise time rather than experiencing discontinuous jumps in response to incoming spikes [78], [79], [80].”
Stable Learning in Stochastic Network States (2012, result): “For appropriate parameters, balanced networks of integrate-and-fire neurons can display asynchronous irregular (AI) states (Brunel, 2000; Vogels and Abbott, 2005; Kumar et al., 2008; El Boustani and Destexhe, 2009; Marre et al., 2009) that produce spike discharge patterns similar to those reported…”
Correlations and Neuronal Population Information (2016, background): “…response statistics places strong constraints on network models (van Vreeswijk & Sompolinsky 1996, Shadlen & Newsome 1998, Vogels & Abbott 2005, Vogels et al. 2005), capturing the magnitude and pattern of correlations under different stimulus and behavioral regimes may provide important clues…”
A Diffusive Homeostatic Signal Maintains Neural Heterogeneity and Responsiveness in Cortical Networks (2015, methodology): “Spiking activity remained asynchronous after STDP, as shown by the distribution of inter-spike intervals and the spike autocorrelograms, although STDP caused weakly synchronous activity in networks without any form of homeostasis (Fig 8Bii–8Biv) [20].”
Simulation of networks of spiking neurons: A review of tools and strategies (2006, unknown): “It has the same parameters as above, except that every cell needs to be depolarized by about 10 mV, which was implemented by setting E L = − 49 mV (see Vogels and Abbott 2005).”
Inhibitory synaptic plasticity: spike timing-dependence and putative network function (2013, background): “1 Department of Physiology, Anatomy, and Genetics, Centre for Neural Circuits and Behaviour, University of Oxford, Oxford, UK 2 School of Computer and Communication Sciences and School of Life Sciences, Brain Mind Institute, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland 3 Ne”
Synaptic Plasticity in Neural Networks Needs Homeostasis with a Fast Rate Detector (2013, background): “The networks we study consist of leaky integrate-and-fire neurons with a relative refractory mechanism connected by conductance-based synapses [46].”
The remarkable robustness of surrogate gradient learning for instilling complex function in spiking neural networks (2020, methodology): “We used leaky integrate-and-fire neurons with current-based exponential synapses (Gerstner, Kistler, Naud, & Paninski, 2014; Vogels & Abbott, 2005).”
Balanced Networks of Spiking Neurons with Spatially Dependent Recurrent Connections (2013, result): “These findings are consistent with previous studies showing that balanced networks can exhibit “stable chaos” characterized by exponentially long transients, sensitivity to small but finite perturbations and insensitivity to sufficiently small perturbations [10,36–39].”