TY  - RPRT
TI  - Sparse Training of Neural Networks based on Multilevel Mirror Descent
AU  - Yannick Lunk
AU  - Sebastian J. Scott
AU  - Leon Bungert
PY  - 2026
UR  - https://arxiv.org/abs/2602.03535
ID  - 2602.03535
ER  -