Temporal Explanation Drift: Measuring and analyzing temporal explanation drift in deep neural networks
Computing and Artificial Intelligence · 26.12.2025 · 3 (4)
DOI: 10.59400/cai4258 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computing and Artificial Intelligence, online erschienen am 26.12.2025. Verfasst von Zaryab Rahman und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The evaluation of explainable artificial intelligence (XAI) methods has largely focused on the properties of a single, fully-trained model. This static view overlooks a critical dimension of reliability: the stability of an explanation throughout the training process. In this paper, we introduce a formal framework to measure and analyze Temporal Explanation Drift (TED), quantifying how a model’s explanations for a consistent, correctly classified input change from one training epoch to the next. Our analysis on Vision Transformers reveals a fundamental decoupling of stability: the high-level semantic focus of an explanation stabilizes early, while its fine-grained structural representation continues to drift long after predictive accuracy has converged. We demonstrate that this is a general phenomenon that extends to Convolutional Neural Networks (CNNs), but find that its magnitude is critically dependent on the explanation method. Our experiments reveal that post-hoc, gradient-based methods like Grad-CAM are dramatically less stable than intrinsic methods like attention rollout. A final, controlled experiment scientifically dissects these effects, showing that the choice of explanation method is the primary driver of temporal instability, which is further modulated by a nuanced interaction with the model’s architecture. Our work establishes temporal stability as a distinct and vital property for evaluating the trustworthiness of explanations and provides a practical framework for its measurement.
Zitieren
APA
Rahman, Z., & Ottoborgo, M. (2025). Temporal Explanation Drift: Measuring and analyzing temporal explanation drift in deep neural networks. Computing and Artificial Intelligence, 3(4). https://doi.org/10.59400/cai4258
Vancouver
Rahman Z, Ottoborgo M. Temporal Explanation Drift: Measuring and analyzing temporal explanation drift in deep neural networks. Computing and Artificial Intelligence. 2025;3(4). doi:10.59400/cai4258
BibTeX
@article{Rahman2025,
author = {Rahman, Zaryab and Ottoborgo, Mattia},
title = {Temporal Explanation Drift: Measuring and analyzing temporal explanation drift in deep neural networks},
journal = {Computing and Artificial Intelligence},
year = {2025},
volume = {3},
number = {4},
doi = {10.59400/cai4258}
}
EUVIMED