TY  - RPRT
TI  - Estimating the Brittleness of AI: Safety Integrity Levels and the Need for Testing Out-Of-Distribution Performance
AU  - Andrew J. Lohn
PY  - 2020
UR  - https://arxiv.org/abs/2009.00802
ID  - 2009.00802
ER  -