paper-with-me

홈 › Papers

On Membership Inference Attacks in Knowledge Distillation

2025-05-17 · Ziyao Cui, Minxing Zhang, Jian Pei

Nowadays, Large Language Models (LLMs) are trained on huge datasets, some including sensitive information. This poses a serious privacy concern because privacy attacks such as Membership Inference Attacks (MIAs) may detect this sensitive information. While knowledge distillation compresses LLMs into efficient, smaller student models, its impact on privacy remains underexplored. In this paper, we investigate how knowledge distillation affects model robustness against MIA. We focus on two questions. First, how is private data protected in teacher and student models? Second, how can we strengthen privacy preservation against MIAs in knowledge distillation? Through comprehensive experiments, we show that while teacher and student models achieve similar overall MIA accuracy, teacher models better protect member data, the primary target of MIA, whereas student models better protect non-member data. To address this vulnerability in student models, we propose 5 privacy-preserving distillation methods and demonstrate that they successfully reduce student models' vulnerability to MIA, with ensembling further stabilizing the robustness, offering a reliable approach for distilling more secure and efficient student models. Our implementation source code is available at https://github.com/richardcui18/MIA_in_KD.

📄 PDF Abstract BibTeX arXiv:2505.11837

Code (1)

richardcui18/mia_in_kd 공식 구현 pytorch

Tasks

Knowledge DistillationPrivacy Preserving

Methods 이 논문이 사용한 방법론

Focus 설명 없음
Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

GLiRA: Black-Box Membership Inference Attack via Knowledge Distillation

2024-05-13 · Andrey V. Galichin, Mikhail Pautov, Alexey Zhavoronkin, Oleg Y. Rogov 외

While Deep Neural Networks (DNNs) have demonstrated remarkable performance in tasks related to perception and control, there are still several unresolved concerns regarding the privacy of their training data, particularl…

image-classificationImage ClassificationInference AttackKnowledge Distillation+1

Students Parrot Their Teachers: Membership Inference on Model Distillation

2023-03-06 · NeurIPS 2023 11

Model distillation is frequently proposed as a technique to reduce the privacy leakage of machine learning. These empirical privacy defenses rely on the intuition that distilled ``student'' models protect the privacy of …

Knowledge Distillation

Mitigating Membership Inference Attacks by Self-Distillation Through a Novel Ensemble Architecture

2021-10-15 · Xinyu Tang, Saeed Mahloujifar, Liwei Song, Virat Shejwalkar 외

Membership inference attacks are a key measure to evaluate privacy leakage in machine learning (ML) models. These attacks aim to distinguish training members from non-members by exploiting differential behavior of the mo…

Privacy Preserving

Membership Privacy for Machine Learning Models Through Knowledge Transfer

2019-06-15 · Virat Shejwalkar, Amir Houmansadr

Large capacity machine learning (ML) models are prone to membership inference attacks (MIAs), which aim to infer whether the target sample is a member of the target model's training dataset. The serious privacy concerns …

BIG-bench Machine LearningGeneral ClassificationInference AttackKnowledge Distillation+2

Unveiling Client Privacy Leakage from Public Dataset Usage in Federated Distillation

2025-02-11 · Haonan Shi, Tu Ouyang, An Wang

Federated Distillation (FD) has emerged as a popular federated training framework, enabling clients to collaboratively train models without sharing private data. Public Dataset-Assisted Federated Distillation (PDA-FD), w…

Federated LearningInference Attack