Alexander A. Alemi, Ian Fischer, Joshua V. Dillon
Published 2018-07-02Version 1
We present a simple case study, demonstrating that Variational Information Bottleneck (VIB) can improve a network's classification calibration as well as its ability to detect out-of-distribution data. Without explicitly being designed to do so, VIB gives two natural metrics for handling and quantifying uncertainty.
Comments: 10 pages, 7 figures. Accepted to UAI 2018 - Uncertainty in Deep Learning Workshop
Successor Uncertainties: exploration and uncertainty in temporal difference learning
Uncertainty in Neural Processes
Uncertainty in Gradient Boosting via Ensembles
![QR code for arXiv:1807.00906 [cs.LG]](http://oss.arxitics.com/images/favicon.svg)