On the Plasticity Collapse in Continual Machine Unlearning

2026-08-29

AI Summary

本文研究了持续机器遗忘中的塑性崩溃问题，通过理论分析和实验验证了该现象，并提出需要开发保持塑性的遗忘算法。

Abstract

Machine unlearning enables deep neural networks to selectively remove the influence of specific data in response to privacy and regulatory requirements. While prior work largely studies single-shot unlearning, real-world systems must accommodate continual unlearning, where multiple unlearning requests occur sequentially over time. In this work, we identify a fundamental limitation of this setting: plasticity collapse, a progressive breakdown in a model's ability to effectively forget. Through theoretical analysis of continual unlearning dynamics, we show that continual unlearning operations accumulate geometric constraints in parameter space, leading to saturated subspaces that restrict future updates. This structural effect induces two distinct failure modes: (1) Forward failure -- diminishing forgetting quality for subsequent tasks, and (2) Backward failure -- spontaneous re-memorization of previously forgotten information. Extensive experiments across multiple architectures, datasets, and methods in image classification confirm that plasticity collapse is not an artifact of specific implementations, but a pervasive phenomenon inherent to continual unlearning. Our findings reveal a critical barrier to the long-term reliability of machine unlearning systems and motivate the development of plasticity-preserving unlearning algorithms. Our code is available at https://github.com/TIML-Group/Continual-Machine-Unlearning-Plasticity-Collapse

Problem

• continual unlearning
• plasticity collapse
• forgetting quality
• re-memorization

Innovation

• Continual Machine Unlearning
• Plasticity Collapse
• Geometric Constraints in Parameter Space
• Forward Failure
• Backward Failure