Abstraction as a Memory-Efficient Inductive Bias for Continual Learning

TL;DR

This paper introduces Abstraction-Augmented Training (AAT), a novel memory-efficient method for continual learning that leverages structural abstraction to stabilize learning without replay buffers.

Contribution

AAT is a new loss-level modification that encourages models to learn shared relational structures, reducing memory needs and outperforming replay-based methods in continual learning.

Findings

AAT performs comparably or better than experience replay baselines.

AAT requires no additional memory and minimal training modifications.

Effective on relational and narrative datasets.

Abstract

The real world is non-stationary and infinitely complex, requiring intelligent agents to learn continually without the prohibitive cost of retraining from scratch. While online continual learning offers a framework for this setting, learning new information often interferes with previously acquired knowledge, causes forgetting and degraded generalization. To address this, we propose Abstraction-Augmented Training (AAT), a loss-level modification encouraging models to capture the latent relational structure shared across examples. By jointly optimizing over concrete instances and their abstract representations, AAT introduces a memory-efficient inductive bias that stabilizes learning in strictly online data streams, eliminating the need for a replay buffer. To capture the multi-faceted nature of abstraction, we introduce and evaluate AAT on two benchmarks: a controlled relational dataset where abstraction is realized through entity masking, and a narrative dataset where abstraction is expressed through shared proverbs. Our results show that AAT achieves performance comparable to or exceeding strong experience replay (ER) baselines, despite requiring zero additional memory and only minimal changes to the training objective. This work highlights structural abstraction as a powerful, memory-free alternative to ER.