A recent study (Rice et al., 2020) revealed overfitting to be a dominant phenomenon in
adversarially robust training of deep networks, and that appropriate early-stopping of …

Pruning large neural networks to create high-quality, independently trainable sparse masks,
which can maintain similar performance to their dense counterparts, is very desirable due to …

Flatness of the loss curve is conjectured to be connected to the generalization ability of
machine learning models, in particular neural networks. While it has been empirically …

The intuition that local flatness of the loss landscape is correlated with better generalization
for deep neural networks (DNNs) has been explored for decades, spawning many different …

Current approximate posteriors in Bayesian neural networks (BNNs) exhibit a crucial
limitation: they fail to maintain invariance under reparameterization, ie BNNs assign different …

Despite extensive research on adversarial examples, the underlying mechanisms of
adversarially robust generalization, a critical yet challenging task for deep learning, remain …

This thesis investigates state-of-the-art methods and underlying theories in supervised deep
learning, aiming to apply insights from these techniques to broader inverse problem settings …

Deep learning, the study of multi-layered artificial neural networks, has received tremendous
attention over the course of the last few years. Neural networks are now able to outperform …

Many studies show a positive correlation between the generalization ability of a deep neural
network and the flatness of the minima in its loss landscape. Inspired by this statement …