AI RESEARCH

Muon Converges under Heavy-Tailed Noise: Nonconvex H\"{o}lder-Smooth Empirical Risk Minimization

arXiv CS.LG

ArXi:2603.15059v1 Announce Type: new Muon is a recently proposed optimizer that enforces orthogonality in parameter updates by projecting gradients onto the Stiefel manifold, leading to stable and efficient