Revisiting Neural Activation Coverage for Uncertainty Estimation

arXiv:2604.22360v1 Announce Type: new 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.

Related

• eCP: Equivariant Conformal Prediction with pre-trained models
• Uncertainty Quantification in PINNs for Turbulent Flows: Bayesian Inference and Repulsive Ensembles
• Cosmological Analysis with Calibrated Neural Quantile Estimation and Approximate Simulators
• Bayesian Neural Networks: An Introduction and Survey

Source: arXiv cs.LG | 2026-04-27