TY  - RPRT
TI  - Neural networks trained with SGD learn distributions of increasing complexity
AU  - Maria Refinetti
AU  - Alessandro Ingrosso
AU  - Sebastian Goldt
PY  - 2023
UR  - https://arxiv.org/abs/2211.11567
ID  - 2211.11567
ER  -