Decoupling Generalizability and Membership Privacy Risks in Neural Networks
Not on the map. Why?
Abstract (arXiv)
A deep learning model usually has to sacrifice some utilities when it acquires some other abilities or characteristics. Privacy preservation has such trade-off relationships with utilities. The loss disparity between various defense approaches implies the potential to decouple generalizability and privacy risks to maximize privacy gain. In this paper, we identify that the model's generalization and privacy risks exist in different regions in deep neural network architectures. Based on the observations that we investigate, we propose Privacy-Preserving Training Principle (PPTP) to protect model components from privacy risks while minimizing the loss in generalizability. Through extensive evaluations, our approach shows significantly better maintenance in model generalizability while enhancing privacy preservation.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 43 of its 59 references that are in the map. First posted on arXiv 2 Feb 2026.
Builds on
43 of 59 references.
References · 43
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
- 2014Explaining and Harnessing Adversarial Examplescited by 7,287
- 2016Deep Learning with Differential Privacycited by 2,682
- 2019A Survey on Bias and Fairness in Machine Learningcited by 1,911
- 2019Machine Unlearningcited by 1,020
- 2017A Closer Look at Memorization in Deep Networkscited by 936
- 2021Membership Inference Attacks From First Principlescited by 839
- 2019Adversarial Training for Free!cited by 638
- 2016Stealing Machine Learning Models via Prediction APIscited by 602
- 2019Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillationcited by 491
- 2022Last Layer Re-Training is Sufficient for Robustness to Spurious Correlationscited by 379
31 more
- 2020Label-Only Membership Inference Attackscited by 371
- 2020Systematic Evaluation of Privacy Risks of Machine Learning Modelscited by 273
- 2019MemGuard: Defending against Black-Box Membership Inference Attacks via Adversarial Examplescited by 183
- 2018Machine Learning with Membership Privacy using Adversarial Regularizationcited by 177
- 2020Regularizing Class-wise Predictions via Self-knowledge Distillationcited by 144
- 2021Deep Learning Through the Lens of Example Difficultycited by 139
- 2019Privacy Risks of Securing Machine Learning Models against Adversarial Examplescited by 136
- 2022The Privacy Onion Effect: Memorization is Relativecited by 128
- 2021On the geometry of generalization and memorization in deep neural networkscited by 71
- 2021Mitigating Membership Inference Attacks by Self-Distillation Through a Novel Ensemble Architecturecited by 63
- 2022Pruning has a disparate impact on model accuracycited by 43
- 2022Membership Inference Attacks and Defenses in Neural Network Pruningcited by 42
- 2022RelaxLoss: Defending Membership Inference Attacks without Losing Utilitycited by 35
- 2020Against Membership Inference Attack: Pruning is All You Needcited by 32
- 2023Overconfidence is a Dangerous Thing: Mitigating Membership Inference Attacks by Enforcing Less Confident Predictioncited by 26
- 2020Privacy Analysis of Deep Learning in the Wild: Membership Inference Attacks against Transfer Learningcited by 19
- 2022A Blessing of Dimensionality in Membership Inference through Regularizationcited by 17
- 2022Leveraging Adversarial Examples to Quantify Membership Information Leakagecited by 16
- 2022Purifier: Defending Data Inference Attacks via Transforming Confidence Scorescited by 15
- 2018Virtual Class Enhanced Discriminative Embedding Learningcited by 13
- 2025Does Training with Synthetic Data Truly Protect Privacy?cited by 12
- 2023MIST: Defending Against Membership Inference Attacks Through Membership-Invariant Subspace Trainingcited by 11
- 2019GAN-Leaks: A Taxonomy of Membership Inference Attacks against Generative Modelscited by 10
- 2022How Does a Deep Learning Model Architecture Impact Its Privacy? A Comprehensive Study of Privacy Attacks on CNNs and Transformerscited by 8
- 2025Severing Spurious Correlations with Data Pruningcited by 7
- 2022Active Token Mixercited by 6
- 2024Mitigating Privacy Risk in Membership Inference by Convex-Concave Losscited by 5
- 2025Rethinking Membership Inference Attacks Against Transfer Learningcited by 5
- 2024Center-Based Relaxed Learning Against Membership Inference Attackscited by 4
- 2024Representation Magnitude has a Liability to Privacy Vulnerabilitycited by 3
- 2025Explaining How Quantization Disparately Skews a Modelcited by 2
Built on by
Citing papers per year: 2026: 1.
1 paper.