Revisiting Neural Activation Coverage for Uncertainty Estimation

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.

Published in 34th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, ESANN 2026