TY  - RPRT
TI  - Langevin Monte-Carlo Provably Learns Depth Two Neural Nets at Any Size and Data
AU  - Dibyakanti Kumar
AU  - Samyak Jha
AU  - Anirbit Mukherjee
PY  - 2025
UR  - https://arxiv.org/abs/2503.10428
ID  - 2503.10428
ER  -