Saved in:
Bibliographic Details
Main Authors: Kim, Gyu Yeol, Oh, Min-hwan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.19156
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915757585072128
author Kim, Gyu Yeol
Oh, Min-hwan
author_facet Kim, Gyu Yeol
Oh, Min-hwan
contents We analyze Muon as originally proposed and used in practice -- using the momentum orthogonalization with a few Newton-Schulz steps. The prior theoretical results replace this key step in Muon with an exact SVD-based polar factor. We prove that Muon with Newton-Schulz converges to a stationary point at the same rate as the SVD-polar idealization, up to a constant factor for a given number $q$ of Newton-Schulz steps. We further analyze this constant factor and prove that it converges to 1 doubly exponentially in $q$ and improves with the degree of the polynomial used in Newton-Schulz for approximating the orthogonalization direction. We also prove that Muon removes the typical square-root-of-rank loss compared to its vector-based counterpart, SGD with momentum. Our results explain why Muon with a few low-degree Newton-Schulz steps matches exact-polar (SVD) behavior at a much faster wall-clock time and explain how much momentum matrix orthogonalization via Newton-Schulz benefits over the vector-based optimizer. Overall, our theory justifies the practical Newton-Schulz design of Muon, narrowing its practice-theory gap.
format Preprint
id arxiv_https___arxiv_org_abs_2601_19156
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Convergence of Muon with Newton-Schulz
Kim, Gyu Yeol
Oh, Min-hwan
Machine Learning
Optimization and Control
We analyze Muon as originally proposed and used in practice -- using the momentum orthogonalization with a few Newton-Schulz steps. The prior theoretical results replace this key step in Muon with an exact SVD-based polar factor. We prove that Muon with Newton-Schulz converges to a stationary point at the same rate as the SVD-polar idealization, up to a constant factor for a given number $q$ of Newton-Schulz steps. We further analyze this constant factor and prove that it converges to 1 doubly exponentially in $q$ and improves with the degree of the polynomial used in Newton-Schulz for approximating the orthogonalization direction. We also prove that Muon removes the typical square-root-of-rank loss compared to its vector-based counterpart, SGD with momentum. Our results explain why Muon with a few low-degree Newton-Schulz steps matches exact-polar (SVD) behavior at a much faster wall-clock time and explain how much momentum matrix orthogonalization via Newton-Schulz benefits over the vector-based optimizer. Overall, our theory justifies the practical Newton-Schulz design of Muon, narrowing its practice-theory gap.
title Convergence of Muon with Newton-Schulz
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2601.19156