PaperScope
LIVE · 2026-09-11 05:40 UTC

Counterfactual Marginalisation: Framework for Evaluating Robustness to Nuisance Variables

Yasin Ibrahim, Hermione Warr, Robin J. Evans, Konstantinos Kamnitsas

Latestcs.CLcs.LGcs.AIcs.CV
arXiv ID
2609.10778 v1
Category
Submitted
2026-09-09

Abstract

Machine learning models can achieve strong test performance while relying on demographic or acquisition-related shortcuts. We propose counterfactual (CF) marginalisation as a test-time evaluation procedure for assessing robustness of classification models to such variables. Given a CF image generator, we intervene on nuisance parent variables such as age or sex, generate CF versions of each test image, and average predictions over a target intervention distribution. This produces intervention-aware predictions that marginalise demographic effects while preserving patient-specific latent information. We use these predictions to define metrics for CF risk, calibration, stability and worst-case sensitivity. We demonstrate this framework's utility for quantitative robustness evaluation.

Comment: Accepted at UNSURE Workshop, MICCAI 2026

arXiv abs page · PDF · same-day batch