mlsanityData to 14 Oct 2026

PEG-Tab: Sampling-Time Record Repair and Release Control for Tabular Synthesis

arXiv 2609.39630 ↗First posted 30 Sep 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Pretrained tabular generators can reproduce training records even when aggregate utility remains high. When retraining is unavailable or too costly, sampling and release are the remaining intervention points. We present PEG-Tab (Post-Training Energy Guidance for Tabular Synthesis), a post-training repair and release-control framework for frozen tabular generators. For each generated row, a generator-native operator creates two alternatives. A shared calibrated score compares the three candidates, favours lower-risk records, and applies a final release check. We instantiate this interface for GReaT, CTGAN, TVAE, and TabDDPM without updating their parameters. Across five datasets and four generator families, PEG-Tab reduces mean Near Copy from 0.078 to 0.027 and lowers aggregate Exact Copy to zero. Relative to a 3× post hoc filter, it retains higher utility in 12 of 16 transfer settings and Pareto-dominates the filter in eight. Gains are concentrated in copy and proximity-related risks.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 15 of its 20 references that are in the map. First posted on arXiv 30 Sep 2026.

Builds on

15 of 20 references.

References · 15

  1. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  2. 2018The Secret Sharer: Evaluating and Testing Unintended Memorization in Neural Networkscited by 864
  3. 2019Modeling Tabular data using Conditional GANcited by 699
  4. 2022TabDDPM: Modelling Tabular Data with Diffusion Modelscited by 377
  5. 2023How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacycited by 153
  6. 2021Winning the NIST Contest: A scalable and general approach to differentially private synthetic datacited by 102
  7. 2022AIM: An Adaptive and Iterative Mechanism for Differentially Private Synthetic Datacited by 90
  8. 2023Membership Inference Attacks against Synthetic Data through Overfitting Detectioncited by 66
  9. 2024Differentially Private Tabular Data Synthesis using Large Language Modelscited by 16
  10. 2025Benchmarking Differentially Private Tabular Data Synthesiscited by 14
  11. 2024Data Plagiarism Index: Characterizing the Privacy Risk of Data-Copying in Tabular Generative Modelscited by 12
  12. 2023Post-processing Private Synthetic Data for Improving Utility on Selected Measurescited by 9
3 more
  1. 2025Synthetic Tabular Data: Methods, Attacks and Defensescited by 9
  2. 2025dpmm: Differentially Private Marginal Models, a Library for Synthetic Tabular Data Generationcited by 3
  3. 2026Differentially Private Synthetic Data via APIs 4: Tabular Datacited by 2

Built on by

Not cited yet.