Kolmogorov-Arnold causal generative models

2026-03-20

AI Summary

This work addresses the challenge of reconciling expressive power with interpretability in deep causal generative models, particularly in high-stakes domains where black-box models hinder auditability. The authors propose a novel causal generative model tailored for mixed-type tabular data, which—uniquely—employs Kolmogorov–Arnold networks to parameterize structural equations. This design yields transparent, directly verifiable causal mechanisms and supports observational, interventional, and counterfactual reasoning. By unifying strong representational capacity with functional interpretability, the approach enables symbolic approximation and causal visualization. Empirical evaluations demonstrate that the model matches state-of-the-art performance on synthetic and semi-synthetic benchmarks while successfully extracting simplified structural equations and interpretable causal effects from real-world cardiovascular data.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Computer Vision: Generative Adversarial Networks (GANs) for Vision

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Causal generative models provide a principled framework for answering observational, interventional, and counterfactual queries from observational data. However, many deep causal models rely on highly expressive architectures with opaque mechanisms, limiting auditability in high-stakes domains. We propose KaCGM, a causal generative model for mixed-type tabular data where each structural equation is parameterized by a Kolmogorov--Arnold Network (KAN). This decomposition enables direct inspection of learned causal mechanisms, including symbolic approximations and visualization of parent--child relationships, while preserving query-agnostic generative semantics. We introduce a validation pipeline based on distributional matching and independence diagnostics of inferred exogenous variables, allowing assessment using observational data alone. Experiments on synthetic and semi-synthetic benchmarks show competitive performance against state-of-the-art methods. A real-world cardiovascular case study further demonstrates the extraction of simplified structural equations and interpretable causal effects. These results suggest that expressive causal generative modeling and functional transparency can be achieved jointly, supporting trustworthy deployment in tabular decision-making settings. Code: https://github.com/aalmodovares/kacgm

Problem

• causal generative models
• interpretability
• Kolmogorov-Arnold Networks
• tabular data
• auditability

Innovation

• Kolmogorov-Arnold Network
• causal generative model
• interpretability
• structural equation
• tabular data