Input-driven circuit reconfiguration in critical recurrent neural networks

TL;DR

This paper introduces a simple recurrent neural network that dynamically reconfigures its signal pathways based solely on input signals, mimicking cortical reconfiguration and enabling targeted signal propagation without changing synaptic weights.

Contribution

It presents a novel single-layer recurrent network that achieves input-driven reconfiguration of pathways using critical dynamics and convolution kernels, without modifying weights.

Findings

Network propagates signals selectively based on input frequencies.

It solves the connectedness problem by controlling wave propagation.

Reconfiguration occurs dynamically without weight changes.

Abstract

Changing a circuit dynamically, without actually changing the hardware itself, is called reconfiguration, and is of great importance due to its manifold technological applications. Circuit reconfiguration appears to be a feature of the cerebral cortex, and hence understanding the neuroarchitectural and dynamical features underlying self-reconfiguration may prove key to elucidate brain function. We present a very simple single-layer recurrent network, whose signal pathways can be reconfigured "on the fly" using only its inputs, with no changes to its synaptic weights. We use the low spatio-temporal frequencies of the input to landscape the ongoing activity, which in turn permits or denies the propagation of traveling waves. This mechanism uses the inherent properties of dynamically-critical systems, which we guarantee through unitary convolution kernels. We show this network solves the classical connectedness problem, by allowing signal propagation only along the regions to be evaluated for connectedness and forbidding it elsewhere.