PulseAugur
中
实时 13:06:44
English(EN) When and How Does Muon Precondition? The Role of Structured Gradient Noise

分析Muon优化器在LLM训练中的优势

研究人员调查了常用于大型语言模型训练的Muon优化器的理论基础。他们的分析表明,在标准的各向同性噪声假设下,Muon除了简单的学习率调整外,没有提供显著的加速。然而,当考虑来自具有不同任务频率的关联记忆中的小批量采样所产生的结构化噪声时,Muon显示出优势。研究表明,Muon可以将任务频率的影响从线性降低到平方根,有助于在学习过程中平衡频繁和稀有任务。此外,Muon还可以避免SignSGD中观察到的由旋转引起的不利迁移,并有效减少采样噪声,同时跟踪梯度变化。 AI

影响 为可能提高LLM训练效率的优化技术提供了理论见解。

排序理由 详细介绍优化技术理论分析的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

分析Muon优化器在LLM训练中的优势

本文如何被排名

Signal score
8 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
详细介绍优化技术理论分析的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, infra
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Same-day
Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.LG TIER_1 English(EN) · Xiaoran Chen, Binghui Li, Liwei Wang, Pinyan Lu ·

    Muon 何时以及如何进行预训练?结构化梯度噪声的作用

    arXiv:2610.31728v1 Announce Type: new Abstract: Muon has become a popular optimizer for large language model training, yet its theoretical foundations remain poorly understood. In this work, we study how gradient noise geometry shapes Muon's preconditioning benefits. We first sho…