Sparse Distributed Memory is a Continual Learner

TL;DR

This paper introduces a biologically inspired Sparse Distributed Memory-based MLP that excels at continual learning without memory replay or task info, offering new training methods for sparse networks.

Contribution

The paper presents a novel biologically inspired MLP variant using SDM that achieves effective continual learning without replay or task labels, with broad applicability.

Findings

Component-wise necessity for continual learning

No reliance on memory replay or task info

Novel training methods for sparse networks

Abstract

Continual learning is a problem for artificial neural networks that their biological counterparts are adept at solving. Building on work using Sparse Distributed Memory (SDM) to connect a core neural circuit with the powerful Transformer model, we create a modified Multi-Layered Perceptron (MLP) that is a strong continual learner. We find that every component of our MLP variant translated from biology is necessary for continual learning. Our solution is also free from any memory replay or task information, and introduces novel methods to train sparse networks that may be broadly applicable.