TY  - RPRT
TI  - Using Pre-Training Can Improve Model Robustness and Uncertainty
AU  - Dan Hendrycks
AU  - Kimin Lee
AU  - Mantas Mazeika
PY  - 2019
UR  - https://arxiv.org/abs/1901.09960
ID  - 1901.09960
ER  -