Probing Privacy Risks in Deep Neural Networks: A Jacobian Rank-Based Analysis of Intermediate Layers
Intelligent Computing · 01.01.2026 · 5
DOI: 10.34133/icomputing.0270 ↗Kurz gefasst
Fachartikel in der Zeitschrift Intelligent Computing, online erschienen am 04.03.2026. Verfasst von Sheng Liu und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 43 Literaturangaben.
Abstract (Originalsprache)
Deep learning models are known to be vulnerable to privacy attacks that can reconstruct or infer sensitive information from the training data. While many privacy assessment methods exist, they are often computationally prohibitive or overlook the considerable information leakage occurring from the model’s intermediate layers. This paper introduces an efficient, attack-independent framework for quantifying privacy risks by analyzing the sensitivity of these internal representations. We discover that the rank of the Jacobian matrix of an intermediate layer’s output with respect to the model’s input serves as a robust proxy for privacy vulnerability. Specifically, we analyze the dynamics of the Jacobian rank throughout training, proposing 2 metrics, rank recovery (RR) and normalized rank recovery (NRR), to specifically quantify a layer’s dynamic transition from generalization to memorization. Through extensive experiments on a diverse range of computer vision models and datasets, including modern residual networks, we demonstrate a strong and consistent positive correlation between our proposed rank-based metrics and the success rates of membership inference attacks. Our findings reveal that deeper layers, which exhibit considerably higher RR, are the primary sources of privacy leakage. This work provides an actionable methodology for developers to identify high-risk layers in real time, facilitating the development of more secure and privacy-preserving machine learning systems without the need for expensive adversarial simulations.
Zitieren
APA
Liu, S., Zhang, X., Huang, T., Zheng, G., Yang, W., & Zhang, J. (2026). Probing Privacy Risks in Deep Neural Networks: A Jacobian Rank-Based Analysis of Intermediate Layers. Intelligent Computing, 5. https://doi.org/10.34133/icomputing.0270
Vancouver
Liu S, Zhang X, Huang T, Zheng G, Yang W, Zhang J. Probing Privacy Risks in Deep Neural Networks: A Jacobian Rank-Based Analysis of Intermediate Layers. Intelligent Computing. 2026;5. doi:10.34133/icomputing.0270
BibTeX
@article{Liu2026,
author = {Liu, Sheng and Zhang, Xuefeng and Huang, Tao and Zheng, Guolong and Yang, Wencheng and Zhang, Ji},
title = {Probing Privacy Risks in Deep Neural Networks: A Jacobian Rank-Based Analysis of Intermediate Layers},
journal = {Intelligent Computing},
year = {2026},
volume = {5},
doi = {10.34133/icomputing.0270}
}
EUVIMED