Short term memory by transient oscillatory dynamics in recurrent neural networks

TL;DR

This paper reveals that short-term memory in recurrent neural networks is encoded in transient oscillatory dynamics, with robustness achieved through contraction onto a slow manifold, offering insights into neural information processing.

Contribution

The study uncovers a novel mechanism where short-term memory is encoded in transient oscillations and demonstrates its robustness across different neural network models and tasks.

Findings

Memory encoded in transient oscillations rather than stationary activity

Robustness to noise via contraction onto a slow manifold

Applicable across multiple neural network models and tasks

Abstract

Despite the significance of short-term memory in cognitive function, the process of encoding and sustaining the input information in neural activity dynamics remains elusive. Herein, we unveiled the significance of transient neural dynamics to short-term memory. By training recurrent neural networks to short-term memory tasks and analyzing the dynamics, the characteristics of the short-term memory mechanism were obtained in which the input information was encoded in the amplitude of transient oscillations, rather than the stationary neural activities. This transient trajectory was attracted to a slow manifold, which permitted the discarding of irrelevant information. Additionally, we investigated the process by which the dynamics acquire robustness to noise. In this transient oscillation, the robustness to noise was obtained by a strong contraction of the neural states after perturbation onto the manifold. This mechanism works for several neural network models and tasks, which implies its relevance to neural information processing in general.