Memory Retrieved from Single Neurons

TL;DR

This paper explores how individual neurons in a Hebbian neural network can access and represent vector memories, proposing a method distinct from Hopfield networks that involves local spreading activity.

Contribution

It introduces a novel approach for associating vector memories with single neurons via local spreading activity, expanding on previous Hebbian network models.

Findings

Single neurons can be associated with vector memories through local activity spreading.

The proposed method differs from Hopfield models by recruiting neighboring neurons.

Open issues and potential extensions for fragment-based memory retrieval are discussed.

Abstract

The paper examines the problem of accessing a vector memory from a single neuron in a Hebbian neural network. It begins with the review of the author's earlier method, which is different from the Hopfield model in that it recruits neighboring neurons by spreading activity, making it possible for single or group of neurons to become associated with vector memories. Some open issues associated with this approach are identified. It is suggested that fragments that generate stored memories could be associated with single neurons through local spreading activity.