TY  - RPRT
TI  - Incorruptible Neural Networks: Training Models that can Generalize to Large Internal Perturbations
AU  - Philip Jacobson
AU  - Ben Feinberg
AU  - Suhas Kumar
AU  - Sapan Agarwal
AU  - T. Patrick Xiao
AU  - Christopher Bennett
PY  - 2026
UR  - https://arxiv.org/abs/2602.07320
ID  - 2602.07320
ER  -