Improving Perturbation-based Explanations by Understanding the Role of Uncertainty Calibration

NeurIPSSpotlight2025

Authors: Thomas Decker, Volker Tresp, Florian Buettner
Venue: NeurIPS 2025
Track: Spotlight

TL;DR

Perturbation-based explanations are widely utilized to enhance the transparency of machine-learning models in practice…

Opening excerpt from the authors’ abstract. source

Read the paper

OpenReview

Topics

uncertainty

← All NeurIPS 2025 Spotlight papers · Browse the whole archive