Reminiscence Attack on Residuals: Exploiting Approximate Machine Unlearning for Privacy

• URL: http://arxiv.org/abs/2507.20573v1
• Date: Mon, 28 Jul 2025 07:12:12 GMT
• Title: Reminiscence Attack on Residuals: Exploiting Approximate Machine Unlearning for Privacy
• Authors: Yaxin Xiao, Qingqing Ye, Li Hu, Huadi Zheng, Haibo Hu, Zi Liang, Haoyang Li, Yijie Jiao,
• Abstract summary: We show that approximate unlearning algorithms fail to adequately protect the privacy of unlearned data.<n>We propose the Reminiscence Attack (ReA), which amplifies the correlation between residuals and membership privacy.<n>We develop a dual-phase approximate unlearning framework that first eliminates deep-layer unlearned data traces and then enforces convergence stability.
• Score: 18.219835803238837
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine unlearning enables the removal of specific data from ML models to uphold the right to be forgotten. While approximate unlearning algorithms offer efficient alternatives to full retraining, this work reveals that they fail to adequately protect the privacy of unlearned data. In particular, these algorithms introduce implicit residuals which facilitate privacy attacks targeting at unlearned data. We observe that these residuals persist regardless of model architectures, parameters, and unlearning algorithms, exposing a new attack surface beyond conventional output-based leakage. Based on this insight, we propose the Reminiscence Attack (ReA), which amplifies the correlation between residuals and membership privacy through targeted fine-tuning processes. ReA achieves up to 1.90x and 1.12x higher accuracy than prior attacks when inferring class-wise and sample-wise membership, respectively. To mitigate such residual-induced privacy risk, we develop a dual-phase approximate unlearning framework that first eliminates deep-layer unlearned data traces and then enforces convergence stability to prevent models from "pseudo-convergence", where their outputs are similar to retrained models but still preserve unlearned residuals. Our framework works for both classification and generation tasks. Experimental evaluations confirm that our approach maintains high unlearning efficacy, while reducing the adaptive privacy attack accuracy to nearly random guess, at the computational cost of 2-12% of full retraining from scratch.

Related papers

• Lie Detector: Unified Backdoor Detection via Cross-Examination Framework [68.45399098884364]
  We propose a unified backdoor detection framework in the semi-honest setting.<n>Our method achieves superior detection performance, improving accuracy by 5.4%, 1.6%, and 11.9% over SoTA baselines.<n> Notably, it is the first to effectively detect backdoors in multimodal large language models.
  arXiv  Detail & Related papers  (2025-03-21T06:12:06Z)
• Pseudo-Probability Unlearning: Towards Efficient and Privacy-Preserving Machine Unlearning [59.29849532966454]
  We propose PseudoProbability Unlearning (PPU), a novel method that enables models to forget data to adhere to privacy-preserving manner. Our method achieves over 20% improvements in forgetting error compared to the state-of-the-art.
  arXiv  Detail & Related papers  (2024-11-04T21:27:06Z)
• Rewind-to-Delete: Certified Machine Unlearning for Nonconvex Functions [11.955062839855334]
  Machine unlearning algorithms aim to efficiently data from a model without it from scratch.<n> certified machine unlearning is a strong theoretical guarantee based on differential generalization.
  arXiv  Detail & Related papers  (2024-09-15T15:58:08Z)
• Nonlinear Transformations Against Unlearnable Datasets [4.876873339297269]
  Automated scraping stands out as a common method for collecting data in deep learning models without the authorization of data owners. Recent studies have begun to tackle the privacy concerns associated with this data collection method. The data generated by those approaches, called "unlearnable" examples, are prevented "learning" by deep learning models.
  arXiv  Detail & Related papers  (2024-06-05T03:00:47Z)
• Hessian-Free Online Certified Unlearning [8.875278412741695]
  We develop an online unlearning algorithm that achieves near-instantaneous data removal.<n>We prove that our proposed method outperforms the state-of-the-art methods in terms of the unlearning and generalization guarantees.
  arXiv  Detail & Related papers  (2024-04-02T07:54:18Z)
• Re-thinking Data Availablity Attacks Against Deep Neural Networks [53.64624167867274]
  In this paper, we re-examine the concept of unlearnable examples and discern that the existing robust error-minimizing noise presents an inaccurate optimization objective. We introduce a novel optimization paradigm that yields improved protection results with reduced computational time requirements.
  arXiv  Detail & Related papers  (2023-05-18T04:03:51Z)
• Learning to Unlearn: Instance-wise Unlearning for Pre-trained Classifiers [71.70205894168039]
  We consider instance-wise unlearning, of which the goal is to delete information on a set of instances from a pre-trained model. We propose two methods that reduce forgetting on the remaining data: 1) utilizing adversarial examples to overcome forgetting at the representation-level and 2) leveraging weight importance metrics to pinpoint network parameters guilty of propagating unwanted information.
  arXiv  Detail & Related papers  (2023-01-27T07:53:50Z)
• Risk Minimization from Adaptively Collected Data: Guarantees for Supervised and Policy Learning [57.88785630755165]
  Empirical risk minimization (ERM) is the workhorse of machine learning, but its model-agnostic guarantees can fail when we use adaptively collected data. We study a generic importance sampling weighted ERM algorithm for using adaptively collected data to minimize the average of a loss function over a hypothesis class. For policy learning, we provide rate-optimal regret guarantees that close an open gap in the existing literature whenever exploration decays to zero.
  arXiv  Detail & Related papers  (2021-06-03T09:50:13Z)
• Continual Learning for Fake Audio Detection [62.54860236190694]
  This paper proposes detecting fake without forgetting, a continual-learning-based method, to make the model learn new spoofing attacks incrementally. Experiments are conducted on the ASVspoof 2019 dataset.
  arXiv  Detail & Related papers  (2021-04-15T07:57:05Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.