Muon with Nesterov Momentum: Heavy-Tailed Noise and (Randomized) Inexact Polar Decomposition

• URL: http://arxiv.org/abs/2605.06884v1
• Date: Thu, 07 May 2026 19:32:39 GMT
• Title: Muon with Nesterov Momentum: Heavy-Tailed Noise and (Randomized) Inexact Polar Decomposition
• Abstract summary: We develop a theory for inexact polar decomposition in non$ matrix optimization undertailed noise.<n>We analyze a randomized low-rank polar decomposition that is substantially more efficient than fullspace methods.
• Score: 11.258577546422295
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Most first-order optimizers treat matrix-valued parameters as vectors, ignoring the intrinsic geometry of hidden-layer weights in neural networks. Muon addresses this mismatch by updating along the polar factor of a momentum matrix, but its theoretical understanding has lagged behind practice. In particular, practical implementations incorporate Nesterov momentum, compute the polar factor only approximately, and operate with stochastic gradients that may be heavy-tailed. We close this gap by developing a convergence theory for Muon with Nesterov momentum and inexact polar decomposition in non-convex matrix optimization under heavy-tailed noise. Our analysis builds on a unified framework for inexact polar decomposition that captures practical iterative approximations such as Newton-Schulz and quantifies how their errors propagate through the optimization dynamics. Under this framework, we establish an optimal iteration and sample complexity of $O \left(\varepsilon^{\frac{-(3α-2)}{(α-1)}} \right)$ for finding an $\varepsilon$-stationary point, where $α\in(1,2]$ denotes the heavy-tail index. For the inexact-polar setting with $σ_1=0$, we also provide guarantees that do not require prior knowledge of $α$. We analyze a randomized low-rank polar decomposition that is substantially more efficient than full-space methods while remaining compatible with our theory. Numerical experiments further demonstrate the effectiveness of the proposed inexact and randomized variants.

Related papers

• Provably Adaptive Linear Approximation for the Shapley Value and Beyond [73.0940890296463]
  A fundamental and long-standing challenge is their efficient approximation.<n>We develop a linear-space algorithm that requires $P(|hatboldsymbol-boldsymbol|_2geq)leq $ for all commonly used semi-values.<n>Our algorithm allows explicit minimization of the mean square error for each specific utility function.
  arXiv  Detail & Related papers  (2026-04-09T16:38:14Z)
• Lyapunov Stability of Stochastic Vector Optimization: Theory and Numerical Implementation [0.0]
  We use a drift--diffusion model for unconstrained vector optimization in which the drift is induced by a common descent direction.<n>We implement the resulting algorithm as an open-source Python framework for multi-objective optimization.<n> Empirical results on DTLZ2 with objective counts from three to fifteen indicate a consistent trade-off.
  arXiv  Detail & Related papers  (2026-03-04T14:04:24Z)
• Regularized Online RLHF with Generalized Bilinear Preferences [68.44113000390544]
  We consider the problem of contextual online RLHF with general preferences.<n>We adopt the Generalized Bilinear Preference Model to capture preferences via low-rank, skew-symmetric matrices.<n>We prove that the dual gap of the greedy policy is bounded by the square of the estimation error.
  arXiv  Detail & Related papers  (2026-02-26T15:27:53Z)
• Optimal Unconstrained Self-Distillation in Ridge Regression: Strict Improvements, Precise Asymptotics, and One-Shot Tuning [61.07540493350384]
  Self-distillation (SD) is the process of retraining a student on a mixture of ground-truth and the teacher's own predictions.<n>We show that for any prediction risk, the optimally mixed student improves upon the ridge teacher for every regularization level.<n>We propose a consistent one-shot tuning method to estimate $star$ without grid search, sample splitting, or refitting.
  arXiv  Detail & Related papers  (2026-02-19T17:21:15Z)
• Why is Normalization Preferred? A Worst-Case Complexity Theory for Stochastically Preconditioned SGD under Heavy-Tailed Noise [17.899443444882888]
  We develop a worst-case complexity theory for inequalityally preconditioned gradient descent (SPSGD)<n>We demonstrate that normalization guarantees convergence to a first-order stationary point at rate $mathcalO(T-fracp-13p-2)$ when problem parameters are known, and $mathcalO(T-fracp-12p)$ when problem parameters are unknown.<n>In contrast, we prove that clipping may fail to converge in the worst case due to the statistical dependence between the preconditioner and the gradient estimates.
  arXiv  Detail & Related papers  (2026-02-13T19:29:17Z)
• RanSOM: Second-Order Momentum with Randomized Scaling for Constrained and Unconstrained Optimization [1.3537117504260623]
  Momentum methods, such as Polyak's Heavy Ball, are the standard for training deep networks but suffer from curvature-induced bias in settings.<n>We propose textbfRanSOM, a unified framework that eliminates this bias by replacing deterministic step sizes with randomized steps drawn from distributions with mean $_t$.<n>We instantiate this framework in two algorithms: textbfRanSOM-E for unconstrained optimization and textbfRanSOM-B for constrained optimization.
  arXiv  Detail & Related papers  (2026-02-06T16:09:36Z)
• Efficiently Solving Discounted MDPs with Predictions on Transition Matrices [6.199300239433395]
  We study Discounted Markov Decision Processes (DMDPs) under a generative model.<n>We propose a novel framework to investigate how a prediction on the transition matrix can enhance the sample efficiency in solving DMDPs.
  arXiv  Detail & Related papers  (2025-02-21T09:59:46Z)
• A stochastic first-order method with multi-extrapolated momentum for highly smooth unconstrained optimization [3.8919212824749296]
  We show that the proposed SFOM can accelerate optimization by exploiting the high-order smoothness of the objective function $f$.<n>To the best of our knowledge, this is the first SFOM to leverage arbitrary-order smoothness of the objective function for acceleration.
  arXiv  Detail & Related papers  (2024-12-19T03:22:47Z)
• Accelerated Stochastic Min-Max Optimization Based on Bias-corrected Momentum [30.01198677588252]
  First-order algorithms require at least $mathcalO(varepsilonepsilon-4)$ complexity to find an $varepsilon-stationary point.<n>We introduce novel momentum algorithms utilizing efficient variable complexity.<n>The effectiveness of the method is validated through robust logistic regression using real-world datasets.
  arXiv  Detail & Related papers  (2024-06-18T20:14:52Z)
• Achieving the Asymptotically Optimal Sample Complexity of Offline Reinforcement Learning: A DRO-Based Approach [36.88301225561535]
  offline reinforcement learning aims to learn from pre-collected datasets without active exploration. Existing approaches adopt a pessimistic stance towards uncertainty by penalizing rewards of under-explored state-action pairs to estimate value functions conservatively. We show that the distributionally robust optimization (DRO) based approach can also address these challenges and is asymptotically minimax optimal
  arXiv  Detail & Related papers  (2023-05-22T17:50:18Z)
• Stochastic Inexact Augmented Lagrangian Method for Nonconvex Expectation Constrained Optimization [88.0031283949404]
  Many real-world problems have complicated non functional constraints and use a large number of data points. Our proposed method outperforms an existing method with the previously best-known result.
  arXiv  Detail & Related papers  (2022-12-19T14:48:54Z)
• Optimal Extragradient-Based Bilinearly-Coupled Saddle-Point Optimization [116.89941263390769]
  We consider the smooth convex-concave bilinearly-coupled saddle-point problem, $min_mathbfxmax_mathbfyF(mathbfx) + H(mathbfx,mathbfy)$, where one has access to first-order oracles for $F$, $G$ as well as the bilinear coupling function $H$. We present a emphaccelerated gradient-extragradient (AG-EG) descent-ascent algorithm that combines extragrad
  arXiv  Detail & Related papers  (2022-06-17T06:10:20Z)
• Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization [50.83356836818667]
  gradient Langevin Dynamics is one of the most fundamental algorithms to solve non-eps optimization problems. In this paper, we show two variants of this kind, namely the Variance Reduced Langevin Dynamics and the Recursive Gradient Langevin Dynamics.
  arXiv  Detail & Related papers  (2022-03-30T11:39:00Z)
• Unified Convergence Analysis for Adaptive Optimization with Moving Average Estimator [75.05106948314956]
  We show that an increasing large momentum parameter for the first-order moment is sufficient for adaptive scaling.<n>We also give insights for increasing the momentum in a stagewise manner in accordance with stagewise decreasing step size.
  arXiv  Detail & Related papers  (2021-04-30T08:50:24Z)
• Optimal Robust Linear Regression in Nearly Linear Time [97.11565882347772]
  We study the problem of high-dimensional robust linear regression where a learner is given access to $n$ samples from the generative model $Y = langle X,w* rangle + epsilon$ We propose estimators for this problem under two settings: (i) $X$ is L4-L2 hypercontractive, $mathbbE [XXtop]$ has bounded condition number and $epsilon$ has bounded variance and (ii) $X$ is sub-Gaussian with identity second moment and $epsilon$ is
  arXiv  Detail & Related papers  (2020-07-16T06:44:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.