<ul><li>Uncertainty quantification is crucial in deep learning for safe and reliable decision-making.</li><li>A new method based on variation due to regularization is proposed for uncertainty quantification in large networks.</li><li>The method adjusts the training loss during fine-tuning and reflects confidence in the output based on all layers of the network.</li><li>Experiments show that the proposed method provides rigorous uncertainty estimates and improves uncertainty quantification quality.</li></ul>

Variation Due to Regularization Tractably Recovers Bayesian Deep Learning

Discover more