arXiv cs.AI· Teresa Pui Yee Yong, Win Kent Ong, Chee Seng Chan·· 6 天前AI 评分41
论文：输出遗忘不等于真正遗忘——重训练视角揭示机器遗忘评估的系统性高估
Erased, but Not Gone: Output Forgetting Is Not True Forgetting
AI 导读
一项 arXiv 研究提出用"重训练一致的表示遗忘"来检验机器遗忘（MU），发现标准输出层评估会系统性高估遗忘效果。在多种遗忘方法、数据集和模型上，模型虽在输出层看似已遗忘，但表示空间仍存在结构化残差：遗忘样本上与重训练部分对齐、保留样本上更不一致，残差集中于重训练相关方向，且常超出独立重训练的波动范围，在再学习攻击下存在信息恢复风险。
正文 · 原文
Abstract:Machine unlearning (MU) is commonly judged by output forgetting, such as low forget-set accuracy or reduced membership inference. But if output-level success can coexist with retraining-inconsistent residuals in representation space, what kind of forgetting are current evaluations actually certifying? We study this question through retraining-consistent representation forgetting, using a population of retrained models (i.e., trained from scratch without the forget data) as an operational reference distribution for ideal forgetting. Across multiple unlearning methods, datasets, and models, our theoretical analysis and empirical results show that standard output-level evaluation can systematically overestimate the success of unlearning. Under this stronger lens, current methods often appear forgotten at the output layer while exhibiting a structured mismatch relative to retraining. They partially align with retraining on forget samples, remain more inconsistent on retain samples, and tend to leave residual discrepancy localized along retraining-related directions rather than diffuse in representation space. This structured mismatch is characterized by forget/retain asymmetry, directional mismatch, and concentrated residuals along retraining-related directions. In many cases, the discrepancies exceed the variability of independent retraining and lead to a risk of information recovery during relearning attacks. Overall, the results suggest that current MU is often evaluated for apparent forgetting rather than retraining-consistent forgetting. More broadly, retraining reveals what output forgetting hides.
| Comments: | 25 pages |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2606.25001 [cs.LG] |
| (or arXiv:2606.25001v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2606.25001 arXiv-issued DOI via DataCite |
Submission history
From: Teresa Pui Yee Yong [view email]
[v1]
Tue, 23 Jun 2026 16:23:53 UTC (7,332 KB)
[v2]
Fri, 9 Oct 2026 07:06:19 UTC (6,621 KB)
来源：arXiv cs.AI · arxiv.org