Optimal Subspace Inference for the Laplace Approximation of Bayesian Neural Networks

Open in new window