mlsanityData to 14 Oct 2026

Bounding data reconstruction attacks with the hypothesis testing interpretation of differential privacy

arXiv 2307.03928 ↗First posted 8 Jul 2023cs.CR cs.AIcited by 15

Not on the map. Why?

Abstract (arXiv)

We explore Reconstruction Robustness (ReRo), which was recently proposed as an upper bound on the success of data reconstruction attacks against machine learning models. Previous research has demonstrated that differential privacy (DP) mechanisms also provide ReRo, but so far, only asymptotic Monte Carlo estimates of a tight ReRo bound have been shown. Directly computable ReRo bounds for general DP mechanisms are thus desirable. In this work, we establish a connection between hypothesis testing DP and ReRo and derive closed-form, analytic or numerical ReRo bounds for the Laplace and Gaussian mechanisms and their subsampled variants.

Key facts. Cited by 15 papers in the map (data through 14 Oct 2026). Builds on 21 of its 28 references that are in the map. First posted on arXiv 8 Jul 2023.

Builds on

21 of 28 references.

References · 21

  1. 2016Deep Learning with Differential Privacycited by 2,728
  2. 2020Extracting Training Data from Large Language Modelscited by 2,189
  3. 2021Membership Inference Attacks From First Principlescited by 878
  4. 2023Extracting Training Data from Diffusion Modelscited by 732
  5. 2020Inverting Gradients -- How easy is it to break privacy in federated learning?cited by 712
  6. 2017Renyi Differential Privacycited by 650
  7. 2018Privacy Amplification by Subsampling: Tight Analyses via Couplings and Divergencescited by 242
  8. 2022Unlocking High-Accuracy Differentially Private Image Classification through Scalecited by 224
  9. 2021Numerical Composition of Differential Privacycited by 180
  10. 2022Reconstructing Training Data with Informed Adversariescited by 148
  11. 2019Deep Learning with Gaussian Differential Privacycited by 114
  12. 2021Optimal Accounting of Differential Privacy via Characteristic Functioncited by 86
9 more
  1. 2019Hypothesis Testing Interpretations and Renyi Differential Privacycited by 85
  2. 2022Connect the Dots: Tighter Discrete Approximations of Privacy Loss Distributionscited by 52
  3. 2022Bounding Training Data Reconstruction in Private (Deep) Learningcited by 46
  4. 2023Bounding Training Data Reconstruction in DP-SGDcited by 46
  5. 2022Analyzing Privacy Leakage in Machine Learning via Multiple Hypothesis Testing: A Lesson From Fanocited by 17
  6. 2022Edgeworth Accountant: An Analytical Approach to Differential Privacy Compositioncited by 14
  7. 2020Sharp Composition Bounds for Gaussian Differential Privacy via Edgeworth Expansioncited by 13
  8. 2023A Randomized Approach for Tight Privacy Accountingcited by 11
  9. 2022The Saddle-Point Accountant for Differential Privacycited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 4; 2025: 6; 2026: 4.

15 papers.

Sort:

By area:

  1. 2024Auditing f-Differential Privacy in One Runcited by 30
  2. 2025Unifying Re-Identification, Attribute Inference, and Data Reconstruction Risks in Differential Privacycited by 15
  3. 2024Attack-Aware Noise Calibration for Differential Privacycited by 8
  4. 2025Position: Bridge the Gaps between Machine Unlearning and AI Regulationcited by 7
  5. 2025Gaussian DP for Reporting Differential Privacy Guarantees in Machine Learningcited by 6
  6. 2024Visual Privacy Auditing with Diffusion Modelscited by 2
  7. 2023Reconciling AI Performance and Data Reconstruction Resilience for Medical Imagingcited by 1
  8. 2024From Mean to Extreme: Formal Differential Privacy Bounds on the Success of Real-World Data Reconstruction Attackscited by 1
  9. 2025Covert Attacks on Machine Learning Training in Passively Secure MPCcited by 1
  10. 2025Securing Transfer-Learned Networks with Reverse Homomorphic Encryptioncited by 0
  11. 2025On Reconstructing Training Data From Bayesian Posteriors and Trained Modelscited by 0
  12. 2026Protecting Private Code in IDE Autocomplete using Differential Privacycited by 0
3 more
  1. 2026Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spotscited by 0
  2. 2026A Sharp Transition in Data Reconstruction under Differential Privacycited by 0
  3. 2026From Pixels to Privacy-Sensitive Features: An Information-Theoretic Framework for Reconstruction Boundscited by 0