Uncertainty Gating for Cost-Aware Explainable Artificial Intelligence

TL;DR

This paper introduces epistemic uncertainty as a low-cost indicator of explanation reliability in AI, enabling more efficient and trustworthy post-hoc interpretability methods across various datasets and models.

Contribution

It proposes using epistemic uncertainty to identify unreliable explanations, improving explanation efficiency and fidelity in cost-aware AI interpretability.

Findings

High epistemic uncertainty correlates with unstable explanations.

Epistemic uncertainty distinguishes faithful from unfaithful explanations.

Results generalize from tabular to image classification tasks.

Abstract

Post-hoc explanation methods are widely used to interpret black-box predictions, but their generation is often computationally expensive and their reliability is not guaranteed. We propose epistemic uncertainty as a low-cost proxy for explanation reliability: high epistemic uncertainty identifies regions where the decision boundary is poorly defined and where explanations become unstable and unfaithful. This insight enables two complementary use cases: `improving worst-case explanations' (routing samples to cheap or expensive XAI methods based on expected explanation reliability), and `recalling high-quality explanations' (deferring explanation generation for uncertain samples under constrained budget). Across four tabular datasets, five diverse architectures, and four XAI methods, we observe a strong negative correlation between epistemic uncertainty and explanation stability. Further analysis shows that epistemic uncertainty distinguishes not only stable from unstable explanations, but also faithful from unfaithful ones. Experiments on image classification confirm that our findings generalize beyond tabular data.