We study neural network loss landscapes through the lens of mode connectivity, the
observation that minimizers of neural networks retrieved via training on a dataset are …

An evaluation criterion for safe and trustworthy deep learning is how well the invariances
captured by representations of deep neural networks (DNNs) are shared with humans. We …

Biological and artificial information processing systems form representations of the world
that they can use to categorize, reason, plan, navigate, and make decisions. To what extent …