Id: arxiv/2606.30225 Url: https://arxiv.org/abs/2606.30225 Tags: math.OC Score: 0 Title: Fast Adaptive Tensor Methods Under Local Smoothness Domain: math.OC Source: arxiv Authors: Sadok Jerad Reviews: 0 Subject: math.OC Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-06-29 Page: https://delphi.science/r/arxiv/2606.30225 A new, fast adaptive regularization methods is proposed and analyzed under local Lipschitz smoothness of the $p$-th order tensor. For nonconvex problems, it achieves the optimal $\mathcal{O}\!\left(|\log(ε)|ε^{-(p+1)/p}\right)$ complexity to obtain first-order $ε$-stationary points and in the convex case, it yields $\mathcal{O}\!\left(|\log(ε)|ε^{-1/p}\right)$ iterations to drive the optimality gap below $ε$, thus matching the complexity bounds of standard tensor methods under global Lipschitz smoothness yp to logarithmic terms. The proposed algorithm follows the line of standard tensor…