mlsanityData to 7 Oct 2026

RelaxLoss: Defending Membership Inference Attacks without Losing Utility

arXiv 2207.05801 ↗First posted 12 Jul 2022cs.LG cs.CRcited by 35

On the map: LLM safety and security → Membership inference attacks

Abstract (arXiv)

As a long-term threat to the privacy of training data, membership inference attacks (MIAs) emerge ubiquitously in machine learning models. Existing works evidence strong connection between the distinguishability of the training and testing loss distributions and the model's vulnerability to MIAs. Motivated by existing results, we propose a novel training framework based on a relaxed loss with a more achievable learning target, which leads to narrowed generalization gap and reduced privacy leakage. RelaxLoss is applicable to any classification model with added benefits of easy implementation and negligible overhead. Through extensive evaluations on five datasets with diverse modalities (images, medical data, transaction records), our approach consistently outperforms state-of-the-art defense mechanisms in terms of resilience against MIAs as well as model utility. Our defense is the first that can withstand a wide range of attacks while preserving (or even improving) the target model's utility. Source code is available at https://github.com/DingfanChen/RelaxLoss

Key facts. Cited by 35 papers in the map (data through 7 Oct 2026). Placed in Membership inference attacks, one of 12 directions in LLM safety and security. Builds on 30 of its 48 references that are in the map. First posted on arXiv 12 Jul 2022.

Builds on

30 of 48 references.

Closest ancestors · 2

Same direction, newest first

  1. 2022On the Privacy Properties of GAN-generated Samples3 Jun
  2. 2021Practical Blind Membership Inference Attack via Differential Comparisons5 Jan

Other references · 28

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  4. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  5. 2022Generative Adversarial Networkscited by 5,603
  6. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  7. 2017On Calibration of Modern Neural Networkscited by 4,860
  8. 2016Deep Learning with Differential Privacycited by 2,682
  9. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  10. 2019When Does Label Smoothing Help?cited by 985
  11. 2020DivideMix: Learning with Noisy Labels as Semi-supervised Learningcited by 691
  12. 2019Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillationcited by 491
16 more
  1. 2018ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Modelscited by 475
  2. 2017Regularizing Neural Networks by Penalizing Confident Output Distributionscited by 461
  3. 2020Label-Only Membership Inference Attackscited by 371
  4. 2016Semi-supervised Knowledge Transfer for Deep Learning from Private Training Datacited by 347
  5. 2013The Composition Theorem for Differential Privacycited by 290
  6. 2020Systematic Evaluation of Privacy Risks of Machine Learning Modelscited by 273
  7. 2018Scalable Private Learning with PATEcited by 266
  8. 2019White-box vs Black-box: Bayes Optimal Strategies for Membership Inferencecited by 265
  9. 2019MemGuard: Defending against Black-Box Membership Inference Attacks via Adversarial Examplescited by 183
  10. 2018Machine Learning with Membership Privacy using Adversarial Regularizationcited by 177
  11. 2017Machine Learning Models that Remember Too Muchcited by 165
  12. 2017PacGAN: The power of two samples in generative adversarial networkscited by 104
  13. 2019BackPACK: Packing more into backpropcited by 77
  14. 2015Efficient Per-Example Gradient Computationscited by 42
  15. 2020On the Effectiveness of Regularization Against Membership Inference Attackscited by 22
  16. 2020On the Difficulty of Membership Inference Attackscited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 7; 2024: 16; 2025: 9; 2026: 3.

35 papers.

Sort:

By area:

  1. 2023Membership Inference Attacks against Language Models via Neighbourhood Comparisoncited by 249
  2. 2024Do Membership Inference Attacks Work on Large Language Models?cited by 188
  3. 2023Low-Cost High-Power Membership Inference Attackscited by 99
  4. 2024Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systemscited by 63
  5. 2023Controlling the Extraction of Memorized Data from Large Language Models via Prompt-Tuningcited by 40
  6. 2024Evaluations of Machine Learning Privacy Defenses are Misleadingcited by 35
  7. 2024On the Privacy Risk of In-context Learningcited by 31
  8. 2024Exposing Privacy Gaps: Membership Inference Attack on Preference Data for LLM Alignmentcited by 24
  9. 2023On the Trustworthiness Landscape of State-of-the-art Generative Models: A Survey and Outlookcited by 15
  10. 2025Does Training with Synthetic Data Truly Protect Privacy?cited by 12
  11. 2025Membership Inference Attacks as Privacy Tools: Reliability, Disparity and Ensemblecited by 11
  12. 2024SoK: Reducing the Vulnerability of Fine-tuned Language Models to Membership Inference Attackscited by 10
23 more
  1. 2024Privacy-Preserving Low-Rank Adaptation against Membership Inference Attacks for Latent Diffusion Modelscited by 9
  2. 2025Rectifying Privacy and Efficacy Measurements in Machine Unlearning: A New Inference Attack Perspectivecited by 8
  3. 2024Mitigating Privacy Risk in Membership Inference by Convex-Concave Losscited by 5
  4. 2025Trustworthy AI: Safety, Bias, and Privacy -- A Surveycited by 5
  5. 2023Discriminative Adversarial Privacy: Balancing Accuracy and Membership Privacy in Neural Networkscited by 4
  6. 2023Diffence: Fencing Membership Privacy With Diffusion Modelscited by 4
  7. 2024Center-Based Relaxed Learning Against Membership Inference Attackscited by 4
  8. 2024Defending Membership Inference Attacks via Privacy-aware Sparsity Tuningcited by 4
  9. 2024Representation Magnitude has a Liability to Privacy Vulnerabilitycited by 3
  10. 2023Membership Privacy Risks of Sharpness Aware Minimizationcited by 2
  11. 2025Do Fairness Interventions Come at the Cost of Privacy: Evaluations for Binary Classifierscited by 2
  12. 2025How does Bayesian Sampling help Membership Inference Attacks?cited by 2
  13. 2024CLMIA: Membership Inference Attacks via Unsupervised Contrastive Learningcited by 1
  14. 2025Estimating Model-Level Membership Inference Vulnerability Without Reference Modelscited by 1
  15. 2025Leverage Unlearning to Sanitize LLMscited by 1
  16. 2026Decoupling Generalizability and Membership Privacy Risks in Neural Networkscited by 1
  17. 2024Amplifying Training Data Exposure through Fine-Tuning with Pseudo-Labeled Membershipscited by 0
  18. 2024Better Membership Inference Privacy Measurement through Discrepancycited by 0
  19. 2024Graph Transductive Defense: a Two-Stage Defense for Graph Membership Inference Attackscited by 0
  20. 2024Resilience and Security of Deep Neural Networks Against Intentional and Unintentional Perturbations: Survey and Research Challengescited by 0
  21. 2025Towards the Anonymization of the Language Modelingcited by 0
  22. 2026Learnability and Privacy Vulnerability are Entangled in a Few Critical Weightscited by 0
  23. 2026Calibrating One-Round Membership Inference with Neighborscited by 0