arXiv AI
Jul 28
Scale Weight Decay and Train Better
arXiv:2607. 23777v1 Announce Type: cross Abstract: The discovery of scaling laws has motivated training neural networks on ever increasing quantities of data.
Convergence Bound and Critical Batch Size of Muon Optimizer
arXiv:2507. 01598v5 Announce Type: replace Abstract: Muon, a recently proposed optimizer that leverages the inherent matrix structure of neural network parameters, has demonstrated strong empirical performance, indicating its potential as a successor to standard optimizers such as AdamW.
Related stories
arXiv AI
Aug 25
A Physical Response-and-Memory Model for Muon Optimization
The paper introduces a physical response-and-memory model for the Muon optimizer, explaining its semi‑orthogonalized momentum update as the maximally dissipative direction under an output‑side safety budget. It treats the weight matrix as a responsive medium with internal stress, showing that momentum corresponds to accumulated stress whose relaxation occurs over multiple timescales—fast and slow. Based on this, the authors propose the Bi‑Maxwell optimizer, which uses a two‑timescale memory kernel and achieves target loss in fewer steps on a public large‑language‑model benchmark.
arXiv Machine Learning
Aug 27
Spectral Allocation: Why Muon Outperforms Adam, and How to Improve Muon
The paper investigates why the orthogonal optimiser Muon outperforms Adam in large language model pretraining by analysing the spectral properties of Transformer loss landscapes. It finds that Muon’s momentum buffers exhibit an anisotropic spectral profile with a volatile head and a tolerant bulk, enabling larger effective step sizes. Building on this insight, the authors propose Spectral‑Aware Muon (SAMuon) and a lightweight variant, which adjust the bulk scaling while keeping the head unchanged, achieving 13–24 % fewer training tokens than Muon without extra FLOPs.
arXiv Machine Learning
Sep 7
Optimizer Memory Schedules for Outscaling the Overtraining Axis
The paper studies how different optimizers perform as training duration (overtraining) increases, focusing on matrix‑preconditioned methods (Muon, SOAP) and a momentum‑scheduled method (ADANA) compared to AdamW. Across models ranging from 51M to 253M parameters and overtraining factors up to 256×, the authors find that optimal learning‑rate schedules, weight‑decay coefficients, and memory settings shift with horizon, and that ADANA consistently outperforms AdamW, especially with log‑time weight decay and momentum cooldown. Muon and SOAP maintain roughly constant token‑efficiency advantages, with SOAP potentially improving at the highest overtraining levels.
arXiv Machine Learning
Jun 30
To Use or not to Use Muon: How Simplicity Bias in Optimizers Matters
arXiv:2603. 00742v2 Announce Type: replace Abstract: While Adam has long been the ubiquitous default optimizer for deep neural networks, Muon has recently seen rapid adoption due to its superior training speed.
arXiv AI
Jul 24
SOAP, Muon, and Beyond: Pushing LLM Pretraining Scales
arXiv:2607. 20548v1 Announce Type: cross Abstract: Higher-order optimizers such as Muon and SOAP offer faster convergence than AdamW, but their computational cost and numerical stability challenges have limited adoption at scale.
arXiv Machine Learning
Sep 11
Musec: MomentUm SpEctral Clipping for Stable Muon-type Training
Musec introduces MomentUm SpEctral Clipping, an optimizer-level, architecture‑agnostic technique that replaces Muon’s spectral flattening with selective spectral clipping to stabilize training. By clipping singular values above a threshold while preserving the momentum’s spectral structure, Musec addresses loss spikes and unbounded weight growth without requiring architecture‑specific changes. Soft Musec, an efficient implementation using smooth spectral saturation via coupled Newton‑Schulz iterations, offers convergence guarantees in nonconvex nonsmooth stochastic optimization and empirically improves stability across diverse learning rates and model sizes.
arXiv Machine Learning
Aug 6
On MUON optimization: From non-convergence to an error analysis with Polar Express and the Newton-Schulz polynomial from implementations
arXiv:2608. 04607v1 Announce Type: cross Abstract: Stochastic gradient descent (SGD) optimization methods are the standard instruments for the training of deep neural networks (DNNs).
arXiv Machine Learning
Aug 18
On the Principles Behind Neural Network Optimizers
arXiv:2608. 16760v1 Announce Type: new Abstract: Reliable optimization is central to neural network (NN) training, yet Adam, the default optimizer for modern LLMs, rests on a fragile foundation.
arXiv Machine Learning
Jun 4
Muon in Associative Memory Learning: Training Dynamics and Scaling Laws
arXiv:2602. 05725v3 Announce Type: replace Abstract: Muon updates matrix parameters via the matrix sign of the gradient and has shown strong empirical gains, yet its dynamics and scaling behavior remain unclear in theory.
arXiv Statistics ML
Aug 26
Lions and Muons: Optimization via Stochastic Frank-Wolfe under Heavy-Tailed Noise
arXiv:2506.04192v4 Announce Type: replace-cross Abstract: Stochastic Frank-Wolfe is a classical optimization method for solving constrained optimization problems. On the other hand, recent optimizers...
arXiv Machine Learning
Jun 26
DMuon: Efficient Distributed Muon Training with Near-Adam Overhead
arXiv:2606. 27153v1 Announce Type: cross Abstract: Matrix-orthogonalization-based optimizers, exemplified by Muon, have demonstrated strong convergence behavior across a wide range of modern deep learning workloads.