paper-with-me

Papers

MUON+: Towards More Effective Muon via One Additional Normalization Step for LLM Pre-training

2026-02-25 · Ruijie Zhang, Yequan Zhao, Ziyue Liu, Zhengyang Wang, Yupeng Su, Liyan Tan, Zheng Zhang arxiv

Muon has recently emerged as a strong optimizer for large language model pre-training, orthogonalizing the momentum matrix via Newton--Schulz polar iterations. A natural intuition is that polar iterations, by flattening the singular spectrum to all ones, should also eliminate column- and row-wise norm imbalance in the update. We show that this is not true in practice: practical polar steps can substantially amplify the imbalance. We term this the post-polar imbalanced update problem, and prove that such imbalance tightens the second-order term in a blockwise descent analysis, weakening Muon's per-step descent guarantee. Motivated by this analysis, we propose Muon+, a one-line fix that inserts a single normalization step after polar orthogonalization. Muon+ adds no optimizer state. Across pre-training experiments on GPT and LLaMA models from 60M to 7B parameters, spanning both compute-optimal budgets and extended token-to-parameter ratios up to approximately 200, Muon+ consistently outperforms Muon in terms of training and validation perplexity, leading to significant overall pre-training speedup.

📄 PDF Abstract BibTeX arXiv:2602.21545

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Aurora: A Leverage-Aware Spectral Optimizer

2026-06-26 · Alec Dewulf, Dhruv Pai, Li Yang, Ashley Zhang 외 arxiv

We show that for tall matrix parameters, like projection matrices in the MLP layers, the Muon update can have row norms that are arbitrarily non-uniform. This can lead to a self-reinforcing feedback loop whereby neurons …

Variance-Adaptive Muon: Accelerating LLM Pretraining with NSR-Modulated and Variance-Scaled Momentum

2026-01-21 · Jingru Li, Yibo Fan, Huan Li arxiv

Large Language Models (LLMs) achieve competitive performance across diverse natural language processing (NLP) tasks, yet pretraining is computationally demanding, making optimizer efficiency an important practical consid…

Delving into Muon and Beyond: Deep Analysis and Extensions

2026-02-04 · Xianbiao Qi, Marco Chen, Jiaquan Ye, Yelin He 외 arxiv

The Muon optimizer has recently attracted considerable attention for its strong empirical performance and use of orthogonalized updates on matrix-shaped parameters, yet its underlying mechanisms and relationship to adapt…

MuonEq: Balancing Before Orthogonalization with Lightweight Equilibration

2026-03-30 · Da Chang, Qiankun Shi, Lvgang Zhang, Yu Li 외 arxiv

Orthogonalized-update optimizers such as Muon improve training of matrix-valued parameters, but existing extensions typically either rescale updates after orthogonalization or use heavier whitening-based preconditioners …

NorMuon: Making Muon more efficient and scalable

2025-10-07 · Zichong Li, Liming Liu, Chen Liang, Weizhu Chen 외 arxiv

The choice of optimizer significantly impacts the training efficiency and computational costs of large language models (LLMs). Recently, the Muon optimizer has demonstrated promising results by orthogonalizing parameter …