Saved in:
Bibliographic Details
Main Authors: Kenfack, Patrik, Kahou, Samira Ebrahimi, Aïvodji, Ulrich
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.09503
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918272542179328
author Kenfack, Patrik
Kahou, Samira Ebrahimi
Aïvodji, Ulrich
author_facet Kenfack, Patrik
Kahou, Samira Ebrahimi
Aïvodji, Ulrich
contents Transformer-based tabular foundation models have recently demonstrated promising in-context learning (ICL) performance on structured data, emerging as competitive alternatives to gradient-boosted trees. However, the fairness implications of this new paradigm remain largely unexplored. We present the first investigation of fairness in tabular ICL, evaluating three recently proposed foundation models--TabPFNv2, TabICL, and TabDPT--on multiple benchmark datasets. To mitigate biases, we explore three pre-processing fairness-enhancing methods: correlation removal (decorrelating input features from the sensitive attribute), group-balanced sample selection (ensuring equal representation of protected groups in context examples), and uncertainty-based sample selection (prioritizing context examples with high sensitive-attribute prediction uncertainty). Our experiments show that the uncertainty-based strategy consistently improves group fairness metrics (e.g., demographic parity, equalized odds, and equal opportunity) with minimal impact on predictive accuracy. We release our code to facilitate reproducibility https://github.com/patrikken/Fair-TabICL.
format Preprint
id arxiv_https___arxiv_org_abs_2505_09503
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Towards Fair In-Context Learning with Tabular Foundation Models
Kenfack, Patrik
Kahou, Samira Ebrahimi
Aïvodji, Ulrich
Machine Learning
Transformer-based tabular foundation models have recently demonstrated promising in-context learning (ICL) performance on structured data, emerging as competitive alternatives to gradient-boosted trees. However, the fairness implications of this new paradigm remain largely unexplored. We present the first investigation of fairness in tabular ICL, evaluating three recently proposed foundation models--TabPFNv2, TabICL, and TabDPT--on multiple benchmark datasets. To mitigate biases, we explore three pre-processing fairness-enhancing methods: correlation removal (decorrelating input features from the sensitive attribute), group-balanced sample selection (ensuring equal representation of protected groups in context examples), and uncertainty-based sample selection (prioritizing context examples with high sensitive-attribute prediction uncertainty). Our experiments show that the uncertainty-based strategy consistently improves group fairness metrics (e.g., demographic parity, equalized odds, and equal opportunity) with minimal impact on predictive accuracy. We release our code to facilitate reproducibility https://github.com/patrikken/Fair-TabICL.
title Towards Fair In-Context Learning with Tabular Foundation Models
topic Machine Learning
url https://arxiv.org/abs/2505.09503