Wide Mean-Field Variational Bayesian Neural Networks Ignore the Data

Abstract

Variational inference enables approximate posterior inference of the highly over-parameterized neural networks that are popular in modern machine learning. Unfortunately, such posteriors are known to exhibit various pathological behaviors. We prove that as the number of hidden units in a single-layer Bayesian neural network tends to infinity, the function-space posterior mean under mean-field variational inference actually converges to zero, completely ignoring the data. This is in contrast to the true posterior, which converges to a Gaussian process. Our work provides insight into the over-regularization of the KL divergence in variational inference.

References in corpus (6)

• Auto-Encoding Variational Bayes
• SGDR: Stochastic Gradient Descent with Warm Restarts
• Reconciling modern machine learning practice and the bias-variance trade-off
• 'In-Between' Uncertainty in Bayesian Neural Networks
• Overpruning in Variational Bayesian Neural Networks
• Exact posterior distributions of wide Bayesian neural networks