Oscillations in Spurious States of the Associative Memory Model with Synaptic Depression

TL;DR

This paper investigates how synaptic depression influences spurious states in an associative memory model, revealing that it destabilizes these states and induces oscillations without affecting true memory states.

Contribution

It provides the first detailed analysis of synaptic depression's effect on spurious attractors in neural network models using Monte Carlo simulations.

Findings

Synaptic depression destabilizes spurious states.

It induces periodic oscillations in the network.

Memory states remain unaffected by synaptic depression.

Abstract

The associative memory model is a typical neural network model, which can store discretely distributed fixed-point attractors as memory patterns. When the network stores the memory patterns extensively, however, the model has other attractors besides the memory patterns. These attractors are called spurious memories. Both spurious states and memory states are equilibrium, so there is little difference between their dynamics. Recent physiological experiments have shown that short-term dynamic synapse called synaptic depression decreases its transmission efficacy to postsynaptic neurons according to the activities of presynaptic neurons. Previous studies have shown that synaptic depression induces oscillation in the network and decreases the storage capacity at finite temperature. How synaptic depression affects spurious states, however, is still unclear. We investigate the effect of synaptic depression on spurious states through Monte Carlo simulation. The results demonstrate that synaptic depression does not affect the memory states but mainly destabilizes the spurious states and induces the periodic oscillations.