Author Identifier (ORCID)
Abstract
Online Continual Learning (OCL) aims to incrementally learn from non-stationary data streams in a one-pass setting, facing the dual challenges of catastrophic forgetting and insufficient training. These challenges intensify the stability-plasticity dilemma, where pre-serving old knowledge conflicts with acquiring new information. In this paper, we propose Projection-based Stabilized Attribution Guidance (PSAG), a modular framework that leverages gradient-based attributions as active guidance signals to selectively preserve task-relevant representations. Our framework consists of three complementary mechanisms: (1) Attribution-Guided Feature Modulation (AGFM) that anchors critical features in the representation space; (2) Importance-Aware Loss Reweighting (IALR) that prioritizes informative samples at the loss level; and (3) Manifold-Consistent Projection (MCP) that emphasizes critical feature dimensions within a Riemannian metric space. To address the issue of attribution instability in online continual learning, we introduce a Reliable Reference Model (R-Model) that maintains consistent knowledge through exponential moving average updates. This design prevents feedback loops during attribution computation and enables reliable feature importance estimation. Extensive experiments on Split CIFAR-10, Split CIFAR-100, and Split Mini-ImageNet demonstrate that PSAG achieves consistent improvements over strong baselines, confirming the efficacy of stabilized attribution guidance in resolving the stability-plasticity dilemma.
Keywords
online continual learning, catastrophic forgetting, continual learning, feature attribution, representation learning, machine learning
Document Type
Journal Article
Date of Publication
1-1-2026
E-ISSN
28358856
Volume
2026
Issue
May
Publication Title
Transactions on Machine Learning Research
Publisher
Journal of Machine Learning Research, Inc.
School
School of Science
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Yu, H., Hu, K., Lu, Z., Lu, S. Q., Wang, Z., & He, F. (2026). PSAG: PSAG: Projection-based stabilized attribution guidance for online continual learning. Transactions on Machine Learning Research. Advance online publication. https://doi.org/