arxiv6.0 / 10

Exact Bayesian layers via characteristic functions

William H. W. Thompson, Guillaume St-Onge, Laurent Hébert-Dufresne, Jean-Gabriel Young

Abstract

Bayesian neural networks provide principled uncertainty estimates, but training them with Monte Carlo sampling can be expensive. Sampling-free variational inference algorithms make training Bayesian neural networks feasible, but existing approaches assume a Gaussian output distribution, which can be inaccurate in applied problems. Here, we introduce CF-Prop, an uncertainty propagation method that uses characteristic functions to compute the exact output distribution of a Bayesian ReLU layer, without any sampling, even in non-Gaussian settings. Because CF-Prop propagates characteristic functions, the evidence lower bound---the objective of variational training---can be computed in closed form for a broad class of likelihoods, including the Gaussian, Poisson, and Gamma. This allows to implement an efficient resulting variational inference algorithm, CF-VI. On CIFAR-100, CF-VI achieves the lowest calibration error of any tested method, improving on the next best method by about 35%. On a heavy-tailed insurance claim dataset, CF-VI accurately captures tail uncertainty that is severely underestimated by both other sampling-free methods and Monte Carlo methods.

Research area

ai reliabilityrobustness to domain shiftsuncertainty quantification
Published
19 Nov 2026
Source
arxiv
Org
University of Vermont
View paper
Sign in to read and join the discussion.