mlsanityData to 14 Oct 2026

On Privacy in Data-Space Tabular Diffusion Models: Influential Factors, Attacker Knowledge, and Metrics

arXiv 2605.06835 ↗First posted 7 May 2026cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Tabular data plays an important role in many fields and industries, including those with elevated privacy considerations and risks. As such, there is a rising interest in generating high-quality synthetic proxies for real tabular data as a means of reducing privacy risk and proprietary data exposure. With data-space tabular diffusion models (TDMs) demonstrating leading performance in synthesizing such data, understanding and measuring the privacy risks associated with these models is imperative. Leveraging state-of-the-art membership inference attacks for such TDMs in both black- and white-box settings, this work quantifies the impact of training setup, synthesis choices, and attacker knowledge on privacy leakage. Moreover, the results demonstrate that adversaries need not have perfect knowledge of the training setup, identical data distributions, or massive compute resources to construct successful attacks. Finally, the substantial pitfalls associated with heuristic privacy metrics, such as distance-to-closest record, are identified.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 47 of its 68 references that are in the map. First posted on arXiv 7 May 2026.

Builds on

47 of 68 references.

References · 47

  1. 2019Optuna: A Next-generation Hyperparameter Optimization Frameworkcited by 2,501
  2. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  3. 2016On Large-Batch Training for Deep Learning: Generalization Gap and Sharp Minimacited by 1,352
  4. 2021Membership Inference Attacks From First Principlescited by 878
  5. 2018ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Modelscited by 498
  6. 2022TabDDPM: Modelling Tabular Data with Diffusion Modelscited by 377
  7. 2021How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Modelscited by 196
  8. 2023Mixed-Type Tabular Data Synthesis with Score-based Diffusion in Latent Spacecited by 186
  9. 2024Large Language Models(LLMs) on Tabular Data: Prediction, Generation, and Understanding -- A Surveycited by 159
  10. 2023Are Diffusion Models Vulnerable to Membership Inference Attacks?cited by 136
  11. 2022Differentially Private Diffusion Modelscited by 119
  12. 2020Revisiting Membership Inference Under Realistic Assumptionscited by 112
35 more
  1. 2020Synthetic Data -- Anonymisation Groundhog Daycited by 108
  2. 2023Low-Cost High-Power Membership Inference Attackscited by 107
  3. 2020On the Generalization Benefit of Noise in Stochastic Gradient Descentcited by 82
  4. 2025Why Diffusion Models Don't Memorize: The Role of Implicit Dynamical Regularization in Trainingcited by 81
  5. 2021Improving Question Answering Model Robustness with Synthetic Adversarial Data Generationcited by 79
  6. 2023Membership Inference Attacks against Synthetic Data through Overfitting Detectioncited by 66
  7. 2024TabDiff: a Mixed-type Diffusion Model for Tabular Data Generationcited by 63
  8. 2022SoK: Let the Privacy Games Begin! A Unified Treatment of Data Inference Privacy in Machine Learningcited by 51
  9. 2020Learning Rates as a Function of Batch Size: A Random Matrix Theory Approach to Neural Network Trainingcited by 50
  10. 2020Generative Data Augmentation for Commonsense Reasoningcited by 49
  11. 2018Data augmentation using synthetic data for time series classification with deep residual networkscited by 38
  12. 2020Investigating Membership Inference Attacks under Data Dependenciescited by 32
  13. 2024Efficacy of Synthetic Data as a Benchmarkcited by 30
  14. 2024ClavaDDPM: Multi-relational Data Synthesis with Cluster-guided Diffusion Modelscited by 28
  15. 2025A Comprehensive Survey of Synthetic Tabular Data Generationcited by 28
  16. 2024SynthEval: A Framework for Detailed Utility and Privacy Evaluation of Tabular Synthetic Datacited by 27
  17. 2023A Linear Reconstruction Approach for Attribute Inference Attacks against Synthetic Datacited by 24
  18. 2024"What do you want from theory alone?" Experimenting with Tight Auditing of Differentially Private Synthetic Data Generationcited by 21
  19. 2024PANORAMIA: Privacy Auditing of Machine Learning Models without Retrainingcited by 17
  20. 2019That which we call privatecited by 16
  21. 2025A Survey on Deep Learning Approaches for Tabular Data Generation: Utility, Alignment, Fidelity, Privacy, Diversity, and Beyondcited by 15
  22. 2023SYNAuG: Exploiting Synthetic Data for Data Imbalance Problemscited by 13
  23. 2024Why Does Differential Privacy with Large Epsilon Defend Against Practical Membership Inference Attacks?cited by 12
  24. 2025The DCR Delusion: Measuring the Privacy Risk of Synthetic Datacited by 11
  25. 2025Winning the MIDST Challenge: New Membership Inference Attacks on Diffusion Models for Tabular Data Synthesiscited by 10
  26. 2024Understanding and Mitigating Memorization in Diffusion Models for Tabular Datacited by 7
  27. 2025Improving Performance, Robustness, and Fairness of Radiographic AI Models with Finely-Controllable Synthetic Datacited by 7
  28. 2022Permutation-Invariant Tabular Data Synthesiscited by 6
  29. 2023Revisiting LARS for Large Batch Training Generalization of Neural Networkscited by 5
  30. 2024DP-TLDM: Differentially Private Tabular Latent Diffusion Modelcited by 5
  31. 2025Ensembling Membership Inference Attacks Against Tabular Generative Modelscited by 4
  32. 2025MIA-EPT: Membership Inference Attack via Error Prediction for Tabular Datacited by 4
  33. 2024Learning Differentially Private Diffusion Models via Stochastic Adversarial Distillationcited by 2
  34. 2025Membership Inference over Diffusion-models-based Synthetic Tabular Datacited by 2
  35. 2026MIDST Challenge at SaTML 2025: Membership Inference over Diffusion-models-based Synthetic Tabular datacited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026SoK: Rethinking Membership Inference in Tabular Machine Learning
  2. 2026FERMI: Exploiting Relations for Membership Inference Against Tabular Diffusion Models
  3. 2025Synth-MIA: A Testbed for Auditing Privacy Leakage in Tabular Data Synthesis
  4. 2023Membership Inference of Diffusion Models
  5. 2024Data Plagiarism Index: Characterizing the Privacy Risk of Data-Copying in Tabular Generative Models
  6. 2022An Empirical Study on the Membership Inference Attack against Tabular Data Synthesis Models
  7. 2025When Tables Leak: Attacking String Memorization in LLM-Based Tabular Data Generation
  8. 2025Flow Matching for Tabular Data Synthesis