Can Acceleration in Gradient-Norm Minimization Be Anytime? Sharp Last-Iterate Limits in Smooth Convex Optimization figure
AlphaXiv 中文概览(可滚动查看)