PRIME: A few primitives can boost robustness to common corruptions
What can linearized neural networks actually say about generalization?
A neural anisotropic view of underspecification in deep learning
Neural Anisotropy Directions
Hold me tight! Influence of discriminative features on deep network boundaries
Optimism in the Face of Adversity: Understanding and Improving Deep Learning through Adversarial Robustness
Redundant features can hurt robustness to distribution shift