A Sobering Look at Tabular Data Generation via Probabilistic Circuits

2026-03-24

AI Summary

This work addresses critical flaws in existing evaluation protocols for tabular data generation, which have led to an overestimation of mainstream models’ performance in real-world scenarios. The authors propose leveraging probabilistic circuits—a class of interpretable, computationally efficient generative models capable of natively handling heterogeneous features—and introduce a more rigorous evaluation framework to expose the true limitations of current state-of-the-art methods. Experimental results demonstrate that lightweight probabilistic circuits achieve comparable or superior generation quality to existing approaches across multiple benchmarks, at significantly lower computational cost. These findings challenge the field’s reliance on increasingly complex architectures and reveal that the apparent performance “saturation” is largely an artifact of inadequate evaluation metrics, thereby advocating for a return to more reliable and reproducible assessment standards in tabular generation research.

Technology Category

• Machine Learning: Probabilistic Circuits and Graphical Models
• Natural Language Processing: Generation
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Tabular data is more challenging to generate than text and images, due to its heterogeneous features and much lower sample sizes. On this task, diffusion-based models are the current state-of-the-art (SotA) model class, achieving almost perfect performance on commonly used benchmarks. In this paper, we question the perception of progress for tabular data generation. First, we highlight the limitations of current protocols to evaluate the fidelity of generated data, and advocate for alternative ones. Next, we revisit a simple baseline -- hierarchical mixture models in the form of deep probabilistic circuits (PCs) -- which delivers competitive or superior performance to SotA models for a fraction of the cost. PCs are the generative counterpart of decision forests, and as such can natively handle heterogeneous data as well as deliver tractable probabilistic generation and inference. Finally, in a rigorous empirical analysis we show that the apparent saturation of progress for SotA models is largely due to the use of inadequate metrics. As such, we highlight that there is still much to be done to generate realistic tabular data. Code available at https://github.com/april-tools/tabpc.

Problem

• tabular data generation
• evaluation metrics
• fidelity
• heterogeneous features
• probabilistic models

Innovation

• Probabilistic Circuits
• Tabular Data Generation
• Evaluation Metrics
• Heterogeneous Data
• Generative Modeling