Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learning.

Arsenii Ashukha, Alexander Lyzhov, Dmitry Molchanov, Dmitry P. Vetrov

Browse the full ICLR paper archive.