TY  - RPRT
TI  - Provably Safeguarding a Classifier from OOD and Adversarial Samples: an Extreme Value Theory Approach
AU  - Nicolas Atienza
AU  - Christophe Labreuche
AU  - Johanne Cohen
AU  - Michele Sebag
PY  - 2025
UR  - https://arxiv.org/abs/2501.10202
ID  - 2501.10202
ER  -