Explaining Concept Shift with Interpretable Feature Attribution

ICLR 2026 Conference Submission19832 Authors

Keywords: distribution shift, generalized additive model, knockoffs, explainable machine learning

Abstract:

Regardless the amount of data a machine learning (ML) model is trained on, there will
inevitably be data that differs from their training set, lowering model performance. Concept
shift occurs when the distribution of labels conditioned on the features changes, making
even a well-tuned ML model to have learned a fundamentally incorrect representation.
Identifying these shifted features provides unique insight into how one dataset differs from
another, considering the difference may be across a scientifically relevant dimension, such
as time, disease status, population, etc. In this paper, we propose SGShift, a model for
detecting concept shift in tabular data and attributing reduced model performance to a
sparse set of shifted features. We frame concept shift as a feature selection task to learn
the features that can explain performance differences between models in the source and
target domain. This framework enables SGShift to adapt powerful statistical tools such as
generalized additive models, knockoffs, and absorption towards identifying these shifted
features. We conduct extensive experiments in synthetic and real data across various ML
models and find SGShift can identify shifted features with AUC > 0.9, much higher than
baseline methods, requires few samples in the shifted domain, and is robust in complex
cases of concept shift. Applying SGShift to 2 real world cases in healthcare and genetics
yielded new feature-level explanations of concept shift, including respiratory failure’s
reduced impact on COVID-19 severity after Omicron and European-specific rare variants’
impact on Lupus prevalence.

Primary Area: interpretability and explainable AI

Submission Number: 19832