{"id":"arxiv/2611.17459","url":"https://arxiv.org/abs/2611.17459","tags":["cs.LG","math.OC"],"score":0,"title":"Classification with Large-Step Gradient Descent: Unifying Self-Stabilization and Convergence Rates","domain":"cs.LG","source":"arxiv","authors":["Salma Tarmoun","Lachlan Ewen MacDonald","Konstantinos Emmanouilidis","Ziqing Xu","René Vidal"],"reviews":0,"subject":"cs.LG","summary":"Neural networks are often trained with gradient descent (GD) using step sizes that exceed classical stability thresholds, a regime where standard convergence theory no longer applies. Despite exhibiting non-monotonic loss and oscillatory dynamics, this regime frequently leads to faster convergence and better generalization. Existing theoretical perspectives offer complementary but incomplete insights: one identifies a self-stabilization mechanism that drives GD toward flat regions but does not yield convergence rates; another derives rigorous rates for specific losses but does not explain…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-11-12","reactions":{}}