Revisiting Neural Activation Coverage for Uncertainty Estimation
Machine Learning
2026-04-27 v1
Abstract
Neural activation coverage (NAC) is a recently-proposed technique for out-of-distribution detection and generalization. We build upon this promising foundation and extend the method to work as an uncertainty estimation technique for already-trained artificial neural networks in the domain of regression. Our experiments confirm NAC uncertainty scores to be more meaningful than other techniques, e.g. Monte-Carlo Dropout.
Keywords
Cite
@article{arxiv.2604.22360,
title = {Revisiting Neural Activation Coverage for Uncertainty Estimation},
author = {Benedikt Franke and Nils Förster and Frank Köster and Asja Fischer and Markus Lange and Arne Raulf},
journal= {arXiv preprint arXiv:2604.22360},
year = {2026}
}
Comments
Published in 34th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2026