mlsanityData to 7 Oct 2026

Causally Constrained Data Synthesis for Private Data Release

arXiv 2105.13144 ↗First posted 27 May 2021cs.LG cs.CRcited by 3

On the map: Synthetic data and self-improvement for LLMs → Privacy-preserving tabular data synthesis

Abstract (arXiv)

Making evidence based decisions requires data. However for real-world applications, the privacy of data is critical. Using synthetic data which reflects certain statistical properties of the original data preserves the privacy of the original data. To this end, prior works utilize differentially private data release mechanisms to provide formal privacy guarantees. However, such mechanisms have unacceptable privacy vs. utility trade-offs. We propose incorporating causal information into the training process to favorably modify the aforementioned trade-off. We theoretically prove that generative models trained with additional causal knowledge provide stronger differential privacy guarantees. Empirically, we evaluate our solution comparing different models based on variational auto-encoders (VAEs), and show that causal information improves resilience to membership inference, with improvements in downstream utility.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Placed in Privacy-preserving tabular data synthesis, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 21 of its 49 references that are in the map. First posted on arXiv 27 May 2021.

Builds on

21 of 49 references.

References · 21

  1. 2013Auto-Encoding Variational Bayescited by 3,980
  2. 2016Deep Learning with Differential Privacycited by 2,682
  3. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  4. 2016Variational Inference: A Review for Statisticianscited by 1,466
  5. 2017Advances in Variational Inferencecited by 236
  6. 2018Protection Against Reconstruction and Its Applications in Private Federated Learningcited by 127
  7. 2019Reconstruction and Membership Inference Attacks against Generative Modelscited by 122
  8. 2020DP-CGAN: Differentially Private Synthetic Data and Label Generationcited by 118
  9. 2018EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAEcited by 85
  10. 2020VAEM: a Deep Generative Model for Heterogeneous Mixed Type Datacited by 66
  11. 2020Instructions and Guide for Diagnostic Questions: The NeurIPS 2020 Education Challengecited by 64
  12. 2020A Causal View on Robustness of Neural Networkscited by 53
9 more
  1. 2018Causal Discovery in the Presence of Missing Datacited by 46
  2. 2019PPGAN: Privacy-preserving Generative Adversarial Networkcited by 34
  3. 2019Neuropathic Pain Diagnosis Simulator for Causal Discovery Algorithm Evaluationcited by 32
  4. 2019Generalization in Generative Adversarial Networks: A Novel Perspective from Privacy Protectioncited by 19
  5. 2019Alleviating Privacy Attacks via Causal Learningcited by 19
  6. 2018Answering Range Queries Under Local Differential Privacycited by 12
  7. 2019Differentially Private Synthetic Mixed-Type Data Generation For Unsupervised Learningcited by 11
  8. 2020Hide-and-Seek Privacy Challengecited by 8
  9. 2020Design of a Privacy-Preserving Data Platform for Collaboration Against Human Traffickingcited by 4

Built on by

Cited by year

Citing papers per year: 2023: 3.

3 papers.

  1. 2023Machine Learning for Synthetic Data Generation: A Reviewcited by 118
  2. 2023A Review of the Role of Causality in Developing Trustworthy AI Systemscited by 14
  3. 2023Adversarial Robustness in Unsupervised Machine Learning: A Systematic Reviewcited by 3

Similar, not cited

Close in wording, no citation either way

  1. 2025Does Training with Synthetic Data Truly Protect Privacy?
  2. 2024SafeSynthDP: Leveraging Large Language Models for Privacy-Preserving Synthetic Data Generation Using Differential Privacy
  3. 2025Achieving Hilbert-Schmidt Independence Under R'enyi Differential Privacy for Fair and Private Data Generation
  4. 2023Differentially Private Diffusion Models Generate Useful Synthetic Images
  5. 2026Phantoms and Disclosures: A Statistical Framework for Auditing Privacy in Synthetic Data
  6. 2023On the Inherent Privacy Properties of Discrete Denoising Diffusion Models
  7. 2024Leveraging Programmatically Generated Synthetic Data for Differentially Private Diffusion Training
  8. 2025RewardDS: Privacy-Preserving Fine-Tuning for Large Language Models via Reward Driven Data Synthesis