Reproduce “Improving predictions of Bayesian neural nets via local linearization” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Alexander Immer, Maciej Korzepa, Matthias Bauer
The authors propose a generalized linear model (GLM) predictive using the same Laplace-GGN posterior as Bayesian neural network (BNN) prediction. Linearizing network features around trained parameters avoids reversals in a toy example. On CIFAR-10, AllCNN accuracy is 81.37 ± 0.15% for GLM and 21.74 ± 0.80% for BNN in that setting, not across metrics.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.