On the capacity of a new model of associative memory based on neural cliques

TL;DR

This paper introduces a new associative memory model inspired by neural cliques, demonstrating improved efficiency over Hopfield networks and providing theoretical bounds on its capacity and convergence properties.

Contribution

The paper presents a novel associative memory model based on neural cliques, with proven convergence and capacity bounds, outperforming traditional Hopfield networks.

Findings

Model has efficiency bounded away from zero

Convergence of retrieval dynamics proven

Upper and lower bounds on memory capacity established

Abstract

Based on recent work by Gripon and Berrou, we introduce a new model of an associative memory. We show that this model has an efficiency bounded away from 0 and is therefore significantly more effective than the well known Hopfield model. We prove that the synchronous and asynchronous retrieval dynamics converge and give upper and lower bounds on the memory capacity of the model.