On Bayesian deep learning

Writing

Reasoning about neural networks

Priors, functions, and uncertainty: a language for thinking about neural networks.

Sampling Bayesian neural networks

From a posterior distribution to a collection of plausible neural networks.

Samples as a data modality

Studying collections of networks through their weights, structure, and predictions.

Research

Analyses

Analyses of Bayesian neural network posterior samples.

Sample posterior weights

A Python package to sample from BNN posteriors.

Align posterior weights

A Python package to remove permutation and scaling symmetries from BNN posterior weights.