Online Continual Learning with Maximally Interfered Retrieval

TL;DR

This paper introduces a novel replay sampling method for online continual learning that selects the most interfered samples to improve performance and reduce forgetting, applicable to generative and experience replay settings.

Contribution

It proposes a controlled sampling strategy based on maximal interference, enhancing replay effectiveness in online continual learning.

Findings

Significant performance improvements over random sampling.

Reduced forgetting in continual learning scenarios.

Applicable to both generative and experience replay methods.

Abstract

Continual learning, the setting where a learning agent is faced with a never ending stream of data, continues to be a great challenge for modern machine learning systems. In particular the online or "single-pass through the data" setting has gained attention recently as a natural setting that is difficult to tackle. Methods based on replay, either generative or from a stored memory, have been shown to be effective approaches for continual learning, matching or exceeding the state of the art in a number of standard benchmarks. These approaches typically rely on randomly selecting samples from the replay memory or from a generative model, which is suboptimal. In this work, we consider a controlled sampling of memories for replay. We retrieve the samples which are most interfered, i.e. whose prediction will be most negatively impacted by the foreseen parameters update. We show a formulation for this sampling criterion in both the generative replay and the experience replay setting, producing consistent gains in performance and greatly reduced forgetting. We release an implementation of our method at https://github.com/optimass/Maximally_Interfered_Retrieval.