{"id":"arxiv/2602.15538","url":"https://arxiv.org/abs/2602.15538","tags":["stat.ML","cs.LG","math.OC"],"score":0,"title":"Functional Central Limit Theorem for Stochastic Gradient Descent","domain":"stat.ML","source":"arxiv","authors":["Kessang Flamand","Victor-Emmanuel Brunel"],"reviews":0,"subject":"stat.ML","summary":"We study the asymptotic shape of the trajectory of the stochastic gradient descent algorithm applied to a convex objective function. Under mild regularity assumptions, we prove a functional central limit theorem for the properly rescaled trajectory. Our result characterizes the long-term fluctuations of the algorithm around the minimizer by providing a diffusion limit for the trajectory. In contrast with classical central limit theorems for the last iterate or Polyak-Ruppert averages, this functional result captures the temporal structure of the fluctuations and applies to non-smooth…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-02-17","reactions":{}}