mlsanityData to 7 Oct 2026

CausalWrap: Model-Agnostic Causal Constraint Wrappers for Tabular Synthetic Data

arXiv 2603.02015 ↗First posted 2 Mar 2026cs.LGcited by 2

Not on the map. Why?

Abstract (arXiv)

Tabular synthetic data generators are typically trained to match observational distributions, which can yield high conventional utility (e.g., column correlations, predictive accuracy) yet poor preservation of structural relations relevant to causal analysis and out-of-distribution (OOD) reasoning. When the downstream use of synthetic data involves causal reasoning -- estimating treatment effects, evaluating policies, or testing mediation pathways -- merely matching the observational distribution is insufficient: structural fidelity and treatment-mechanism preservation become essential. We propose CausalWrap (CW), a model-agnostic wrapper that injects partial causal knowledge (PCK) -- trusted edges, forbidden edges, and qualitative/monotonic constraints -- into any pretrained base generator (GAN, VAE, or diffusion model), without requiring access to its internals. CW learns a lightweight, differentiable post-hoc correction map applied to samples from the base generator, optimized with causal penalty terms under an augmented-Lagrangian schedule. We provide theoretical results connecting penalty-based optimization to constraint satisfaction and relating approximate factorization to joint distributional control. We validate CW on simulated structural causal models (SCMs) with known ground-truth interventions, semi-synthetic causal benchmarks (IHDP and an ACIC-style suite), and a real-world ICU cohort (MIMIC-IV) with expert-elicited partial graphs. CW improves causal fidelity across diverse base generators -- e.g., reducing average treatment effect (ATE) error by up to 63% on ACIC and lifting ATE agreement from 0.00 to 0.38 on the intensive care unit (ICU) cohort -- while largely retaining conventional utility.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 24 of its 41 references that are in the map. First posted on arXiv 2 Mar 2026.

Builds on

24 of 41 references.

References · 24

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  3. 2017Counterfactual Fairnesscited by 802
  4. 2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 745
  5. 2019Modeling Tabular data using Conditional GANcited by 695
  6. 2017Causal Effect Inference with Deep Latent-Variable Modelscited by 408
  7. 2022TabDDPM: Modelling Tabular Data with Diffusion Modelscited by 373
  8. 2022Language Models are Realistic Tabular Data Generatorscited by 290
  9. 2016Revisiting Classifier Two-Sample Testscited by 258
  10. 2017Generating Multi-label Discrete Patient Records using Generative Adversarial Networkscited by 197
  11. 2018Data Synthesis based on Generative Adversarial Networkscited by 184
  12. 2017CausalGAN: Learning Causal Implicit Generative Models with Adversarial Trainingcited by 152
12 more
  1. 2017Automated versus do-it-yourself methods for causal inference: Lessons learned from a data analysis competitioncited by 106
  2. 2024How Realistic Is Your Synthetic Data? Constraining Deep Generative Models for Tabular Datacited by 29
  3. 2025A Comprehensive Survey of Synthetic Tabular Data Generationcited by 27
  4. 2018Physics-informed deep generative modelscited by 16
  5. 2024Learning Constrained Optimization with Deep Augmented Lagrangian Methodscited by 16
  6. 2021Causal-TGAN: Generating Tabular Data Using Causal Generative Adversarial Networkscited by 12
  7. 2025TabStruct: Measuring Structural Fidelity of Tabular Datacited by 9
  8. 2025DeCaFlow: A deconfounding causal generative modelcited by 8
  9. 2024Causality for Tabular Data Synthesis: A High-Order Structure Causal Benchmark Frameworkcited by 7
  10. 2025Improving the Generation and Evaluation of Synthetic Data for Downstream Medical Causal Inferencecited by 4
  11. 2023Achieving Constraints in Neural Networks: A Stochastic Augmented Lagrangian Approachcited by 3
  12. 2025Generating Synthetic Relational Tabular Data via Structural Causal Modelscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Workload-Preserving Differentially Private Synthetic Data for Causal Inference via Maximum-Entropy Calibrationcited by 0
  2. 2026Beyond Distributional Fidelity: Causal-Penalized Diffusion for Synthetic Tabular Datacited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Generative Synthetic Data for Causal Inference: Pitfalls, Remedies, and Opportunities
  2. 2026Improving TabPFN's Synthetic Data Generation by Integrating Causal Structure
  3. 2026TabCausal: Pretraining Across Causal Environments for Tabular Causal Discovery
  4. 2026TabSCM: A practical Framework for Generating Realistic Tabular Data
  5. 2025A Technical Exploration of Causal Inference with Hybrid LLM Synthetic Data
  6. 2025CausalDiffTab: Mixed-Type Causal-Aware Diffusion for Tabular Data Generation
  7. 2025CALM: A Causal Analysis Language Model for Tabular Data in Complex Systems with Local Scores, Conditional Independence Tests, and Relation Attributes
  8. 2023Modeling Causal Mechanisms with Diffusion Models for Interventional and Counterfactual Queries