A Comparative Study of Sparse Associative Memories

TL;DR

This paper compares different models of sparse associative memories, analyzing how various parameters like synaptic weights and architectures affect their storage capacity, focusing on patterns with about log N ones.

Contribution

It provides a comparative analysis of models for sparse associative memories, highlighting the impact of design choices on storage capacity.

Findings

Different models exhibit varying storage capacities based on architecture.

Synaptic weight schemes significantly influence retrieval success.

Sparse patterns with log N ones are effectively stored in certain models.

Abstract

We study various models of associative memories with sparse information, i.e. a pattern to be stored is a random string of $0$s and $1$s with about $\log N$ $1$s, only. We compare different synaptic weights, architectures and retrieval mechanisms to shed light on the influence of the various parameters on the storage capacity.