Id: palomar/PALOMAR-2026-11-14-000001 Url: https://github.com/lowrank/last-iterate-sgd-bound/tree/82241483b73d53c8084edb33a7d946d070674562 Json: https://data.palomar-registry.org/entries/PALOMAR-2026-11-14-000001-v1.json Repo: https://github.com/lowrank/last-iterate-sgd-bound Tags: stat.ML, cs.LG, msc:90C15, msc:90C25, msc:68T05 Score: 0 Title: Improved last-iterate bounds for SGD with and without replacement Trust: high Domain: stat.ML Source: palomar Authors: Yimin Zhong Reviews: 0 Subject: stat.ML Category: research Comments: 0 Disputed: false Division: stat Mentions: 0 Verified: true Discussed: 0 Published: 2026-11-14 Formalized: true Formalizes: title=Improved last-iterate bounds for SGD with and without replacement identifier= relationship=formalizes Page: https://delphi.science/r/palomar/PALOMAR-2026-11-14-000001 For a finite family of differentiable convex beta-smooth losses with a population minimizer, this formalization proves an expected last-iterate population-loss bound for stochastic gradient descent under uniform sampling with replacement and uniform sampling without replacement. With delta = min(1, 2 - eta beta), the bound is 280000 times log(T+2)^3 / log(2+log(T+2)) times [R^2 / (delta^3 eta T) + eta sigma_star^2 / delta^4]. It holds for all 0 < eta beta < 2, every positive WR horizon, and every WOR horizon through one full pass, including T = n. Under interpolation at eta = 1/beta, this…