Papers › Causal Pre-training Under the Fairness Lens: An Empirical Study of TabPFN
Causal Pre-training Under the Fairness Lens: An Empirical Study of TabPFN
Qinyi Liu, Mohammad Khalil, Naman Goel
Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).
Foundation models for tabular data, such as the Tabular Prior-data Fitted Network (TabPFN), are pre-trained on a massive number of synthetic datasets generated by structural causal models (SCM). They leverage in-context learning to offer high predictive accuracy in real-world tasks. However, the fairness properties of these foundational models, which incorporate ideas from causal reasoning during pre-training, remain underexplored. In this work, we conduct a comprehensive empirical evaluation of TabPFN and its fine-tuned variants, assessing predictive performance, fairness, and robustness across varying dataset sizes and distributional shifts. Our results reveal that while TabPFN achieves stronger predictive accuracy compared to baselines and exhibits robustness to spurious correlations, improvements in fairness are moderate and inconsistent, particularly under missing-not-at-random (MNAR) covariate shifts. These findings suggest that the causal pre-training in TabPFN is helpful but insufficient for algorithmic fairness, highlighting implications for deploying TabPFN (and similar) models in practice and the need for further fairness interventions.
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Syntology holds the repository link but has not harvested or run code from it.
Results from the paper
The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2601.17912, was issued in January 2026, after that date, so the archive has no leaderboard rows for it.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections