{"id":"arxiv/2606.17364","url":"https://arxiv.org/abs/2606.17364","tags":["math.ST","math.OC","stat.ML"],"score":0,"title":"A Polyak-Ruppert Central Limit Theorem for SA-Adam with Momentum and Non-Convergent Adaptive Preconditioning","domain":"math.ST","source":"arxiv","authors":["Sunyoung An","Xiaoming Huo"],"reviews":0,"subject":"math.ST","summary":"Adaptive optimizers combining preconditioning, momentum, and weight decay (Adam and AdamW) are, under Polyak-Ruppert averaging, candidate engines for one-pass inference. Does the averaged iterate keep the classical Polyak-Ruppert central limit theorem (CLT), with sandwich covariance $H^{-1}SH^{-1}$ (Hessian $H$, gradient covariance $S$), under momentum and non-convergent preconditioning? The preconditioner-only analysis does not carry over: with momentum the canonical decomposition collapses to a tautology. Treating the augmented state (iterate, momentum buffer) as a time-varying linear…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-06-15","reactions":{}}