High-Dimensional Limit of Stochastic Gradient Flow via Dynamical Mean-Field Theory

arXiv:2602.06320v2 Announce Type: replace-cross Abstract: Modern machine learning models are typically trained via multi-pass stochastic gradient descent (SGD) with small batch sizes, and understanding their dynamics in high dimensions is of great interest. However, an analyti...

🔗 Read more: https://arxiv.org/abs/2602.06320

#News #Math #AI #Tech #Neuro #Psychology #Academic