Energy-Based Models for Continual Learning

TL;DR

This paper proposes using Energy-Based Models (EBMs) for continual learning, demonstrating they reduce interference and outperform baselines, while being adaptable to changing data distributions without explicit task boundaries.

Contribution

The paper introduces a simple, efficient EBM approach for continual learning that outperforms existing methods and can enhance other continual learning techniques.

Findings

EBMs outperform baseline methods on several benchmarks.

Contrastive divergence training boosts performance when combined with other methods.

EBMs are adaptable to non-task-specific data distribution changes.

Abstract

We motivate Energy-Based Models (EBMs) as a promising model class for continual learning problems. Instead of tackling continual learning via the use of external memory, growing models, or regularization, EBMs change the underlying training objective to cause less interference with previously learned information. Our proposed version of EBMs for continual learning is simple, efficient, and outperforms baseline methods by a large margin on several benchmarks. Moreover, our proposed contrastive divergence-based training objective can be combined with other continual learning methods, resulting in substantial boosts in their performance. We further show that EBMs are adaptable to a more general continual learning setting where the data distribution changes without the notion of explicitly delineated tasks. These observations point towards EBMs as a useful building block for future continual learning methods.