TY  - RPRT
TI  - Simmering: Sufficient is better than optimal for training neural networks
AU  - Irina Babayan
AU  - Hazhir Aliahmadi
AU  - Greg van Anders
PY  - 2025
DO  - 10.1038/s41467-025-66983-3
UR  - https://arxiv.org/abs/2410.19912
ID  - 2410.19912
ER  -