Membership Inference Attack Should Move On to Distributional Statistics for Distilled Generative Models

• URL: http://arxiv.org/abs/2502.02970v3
• Date: Thu, 19 Jun 2025 06:33:05 GMT
• Title: Membership Inference Attack Should Move On to Distributional Statistics for Distilled Generative Models
• Authors: Muxing Li, Zesheng Ye, Yixuan Li, Andy Song, Guangquan Zhang, Feng Liu,
• Abstract summary: To detect unauthorized data usage in training large-scale generative models, membership inference attacks (MIAs) have proven effective.<n>We find that standard MIAs fail against distilled generative models (i.e., student models) that are increasingly deployed in practice for efficiency.<n>We propose three principles of distribution-based MIAs for detecting unauthorized training data through distilled generative models.
• Score: 31.834967019893227
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: To detect unauthorized data usage in training large-scale generative models (e.g., ChatGPT or Midjourney), membership inference attacks (MIA) have proven effective in distinguishing a single training instance (a member) from a single non-training instance (a non-member). This success is mainly credited to a memorization effect: models tend to perform better on a member than a non-member. However, we find that standard MIAs fail against distilled generative models (i.e., student models) that are increasingly deployed in practice for efficiency (e.g., ChatGPT 4o-mini). Trained exclusively on data generated from a large-scale model (a teacher model), the student model lacks direct exposure to any members (teacher's training data), nullifying the memorization effect that standard MIAs rely on. This finding reveals a serious privacy loophole, where generation-service providers could deploy a student model whose teacher was potentially trained on unauthorized data, yet claim the deployed model is clean because it was not directly trained on such data. Hence, are distilled models inherently unauditable for upstream privacy violations, and should we discard them when we care about privacy? We contend no, as we uncover a memory chain connecting the student and teacher's member data: the distribution of student-generated data aligns more closely with the distribution of the teacher's members than with non-members, thus we can detect unauthorized data usage even when direct instance-level memorization is absent. This leads us to posit that MIAs on distilled generative models should shift from instance-level scores to distribution-level statistics. We further propose three principles of distribution-based MIAs for detecting unauthorized training data through distilled generative models, and validate our position through an exemplar framework. We lastly discuss the implications of our position.

Related papers

• Self-Comparison for Dataset-Level Membership Inference in Large (Vision-)Language Models [73.94175015918059]
  We propose a dataset-level membership inference method based on Self-Comparison. Our method does not require access to ground-truth member data or non-member data in identical distribution.
  arXiv  Detail & Related papers  (2024-10-16T23:05:59Z)
• Detecting Training Data of Large Language Models via Expectation Maximization [62.28028046993391]
  We introduce EM-MIA, a novel membership inference method that iteratively refines membership scores and prefix scores via an expectation-maximization algorithm. EM-MIA achieves state-of-the-art results on WikiMIA.
  arXiv  Detail & Related papers  (2024-10-10T03:31:16Z)
• Blind Baselines Beat Membership Inference Attacks for Foundation Models [24.010279957557252]
  Membership inference (MI) attacks try to determine if a data sample was used to train a machine learning model. For foundation models trained on unknown Web data, MI attacks can be used to detect copyrighted training materials, measure test set contamination, or audit machine unlearning. We show that evaluations of MI attacks for foundation models are flawed, because they sample members and non-members from different distributions.
  arXiv  Detail & Related papers  (2024-06-23T19:40:11Z)
• Do Membership Inference Attacks Work on Large Language Models? [141.2019867466968]
  Membership inference attacks (MIAs) attempt to predict whether a particular datapoint is a member of a target model's training data. We perform a large-scale evaluation of MIAs over a suite of language models trained on the Pile, ranging from 160M to 12B parameters. We find that MIAs barely outperform random guessing for most settings across varying LLM sizes and domains.
  arXiv  Detail & Related papers  (2024-02-12T17:52:05Z)
• Practical Membership Inference Attacks against Fine-tuned Large Language Models via Self-prompt Calibration [32.15773300068426]
  Membership Inference Attacks aim to infer whether a target data record has been utilized for model training. We propose a Membership Inference Attack based on Self-calibrated Probabilistic Variation (SPV-MIA)
  arXiv  Detail & Related papers  (2023-11-10T13:55:05Z)
• Assessing Privacy Risks in Language Models: A Case Study on Summarization Tasks [65.21536453075275]
  We focus on the summarization task and investigate the membership inference (MI) attack. We exploit text similarity and the model's resistance to document modifications as potential MI signals. We discuss several safeguards for training summarization models to protect against MI attacks and discuss the inherent trade-off between privacy and utility.
  arXiv  Detail & Related papers  (2023-10-20T05:44:39Z)
• Membership Inference Attacks against Synthetic Data through Overfitting Detection [84.02632160692995]
  We argue for a realistic MIA setting that assumes the attacker has some knowledge of the underlying data distribution. We propose DOMIAS, a density-based MIA model that aims to infer membership by targeting local overfitting of the generative model.
  arXiv  Detail & Related papers  (2023-02-24T11:27:39Z)
• RelaxLoss: Defending Membership Inference Attacks without Losing Utility [68.48117818874155]
  We propose a novel training framework based on a relaxed loss with a more achievable learning target. RelaxLoss is applicable to any classification model with added benefits of easy implementation and negligible overhead. Our approach consistently outperforms state-of-the-art defense mechanisms in terms of resilience against MIAs.
  arXiv  Detail & Related papers  (2022-07-12T19:34:47Z)
• Decentralized Federated Learning Preserves Model and Data Privacy [77.454688257702]
  We propose a fully decentralized approach, which allows to share knowledge between trained models. Students are trained on the output of their teachers via synthetically generated input data. The results show that an untrained student model, trained on the teachers output reaches comparable F1-scores as the teacher.
  arXiv  Detail & Related papers  (2021-02-01T14:38:54Z)
• Investigating Membership Inference Attacks under Data Dependencies [26.70764798408236]
  Training machine learning models on privacy-sensitive data has opened the door to new attacks that can have serious privacy implications. One such attack, the Membership Inference Attack (MIA), exposes whether or not a particular data point was used to train a model. We evaluate the defence under the restrictive assumption that all members of the training set, as well as non-members, are independent and identically distributed.
  arXiv  Detail & Related papers  (2020-10-23T00:16:46Z)
• Amnesiac Machine Learning [15.680008735220785]
  Recently enacted General Data Protection Regulation affects any data holder that has data on European Union residents. Models are vulnerable to information leaking attacks such as model inversion attacks. We present two data removal methods, namely Unlearning and Amnesiac Unlearning, that enable model owners to protect themselves against such attacks while being compliant with regulations.
  arXiv  Detail & Related papers  (2020-10-21T13:14:17Z)
• Knowledge-Enriched Distributional Model Inversion Attacks [49.43828150561947]
  Model inversion (MI) attacks are aimed at reconstructing training data from model parameters. We present a novel inversion-specific GAN that can better distill knowledge useful for performing attacks on private models from public data. Our experiments show that the combination of these techniques can significantly boost the success rate of the state-of-the-art MI attacks by 150%.
  arXiv  Detail & Related papers  (2020-10-08T16:20:48Z)
• How Does Data Augmentation Affect Privacy in Machine Learning? [94.52721115660626]
  We propose new MI attacks to utilize the information of augmented data. We establish the optimal membership inference when the model is trained with augmented data.
  arXiv  Detail & Related papers  (2020-07-21T02:21:10Z)

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.