Overview

Tabular prediction studies how to learn from rows of examples and columns of mixed numerical, categorical, and sometimes text or relational features. Common approaches include gradient-boosted trees, feature engineering, neural networks, and tabular foundation models pretrained on synthetic data for in-context learning; graph methods can add relationships across rows or tables. Benchmarks compare these methods on classification and regression, while applied work also examines calibration, fairness, interpretability, and deployment costs. No model family dominates across settings: foundation models are often strong on small or in-distribution datasets, while large, high-dimensional, or shifted data still challenge them, making generalization and fair evaluation central open issues.

Latest papers

CardsList

126 papers14 papers in the last four weeks, up 8% on the four weeks before. 0.1% of all new papers.

  1. tensorFM: Low-Rank Approximations of Cross-Order Feature Interactions

    Feb 16, 2026Alessio Mazzetto, Mohammad Mahdi Khalili, Laura Fee Nern +3Feature Interaction ModelingTabular Prediction

  2. TabSieve: Explicit In-Table Evidence Selection for Tabular Prediction

    Feb 12, 2026Yongyao Wang, Ziqi Miao, Lu Yang +4In-Context Example SelectionReinforcement Learning

  3. TabICLv2: A better, faster, scalable, and open tabular foundation model

    Feb 11, 2026Jingang Qu, David Holzmüller, Gaël Varoquaux +1Synthetic Data PretrainingTabular Foundation Models

  4. MedFeat: Model-Aware and Explainability-Driven Feature Engineering with LLMs for Tabular Prediction

    Feb 10, 2026Zizheng Zhang, Yiming Li, Justin Xu +6Interpretable MLClinical Prediction

  5. Grables: When Graph Structure Helps Tabular Prediction

    Feb 3, 2026Tamara Cucumides, Floris GeertsTabular MLTabular Prediction

  6. Strengthening LLMs for Tabular Prediction with Structural Priors

    Oct 20, 2025Pengxiang Cai, Zihao Gao, Wanchen Lian +2Tabular Foundation ModelsRL for Language Models