Zhijie Deng

Accelerated Linearized Laplace Approximation for Bayesian Deep Learning

2022 · arXiv (Cornell University)

Laplace approximation (LA) and its linearized variant (LLA) enable effortless adaptation of pretrained deep neural networks to Bayesian neural networks. The generalized Gauss-Newton (GGN) approximation is typically introduced to improve their tractability. However, LA and …