7 papers
Balancing Learning Rates Across Layers: Exact Two-Step Dynamics and Optimal Scaling in Linear Neural Networks
Tianyu Pang, Vignesh Kothapalli, Shenyang Deng +3
We study optimal learning-rate selection in two-layer and three-layer linear neural networks trained to learn linear target functions. In particular, we derive the exact closed-for…
Depth, Not Data: An Analysis of Hessian Spectral Bifurcation
Shenyang Deng, Boyao Liao, Zhuoli Ouyang +2
The eigenvalue distribution of the Hessian matrix plays a crucial role in understanding the optimization landscape of deep neural networks. Prior work has attributed the well-docum…
HTMuon: Improving Muon via Heavy-Tailed Spectral Correction
Tianyu Pang, Yujie Fang, Zihang Liu +4
Muon has recently shown promising results in LLM training. In this work, we study how to further improve Muon. We argue that Muon's orthogonalized update rule suppresses the emerge…
RMNP: Row-Momentum Normalized Preconditioning for Scalable Matrix-Based Optimization
Shenyang Deng, Zhuoli Ouyang, Tianyu Pang +4
Preconditioned adaptive methods have gained significant attention for training deep neural networks, as they capture rich curvature information of the loss landscape. The central c…
Suspicious Alignment of SGD: A Fine-Grained Step Size Condition Analysis
Shenyang Deng, Boyao Liao, Zhuoli Ouyang +3
This paper explores the suspicious alignment phenomenon in stochastic gradient descent (SGD) under ill-conditioned optimization, where the Hessian spectrum splits into dominant and…
From Spikes to Heavy Tails: Unveiling the Spectral Evolution of Neural Networks
Vignesh Kothapalli, Tianyu Pang, Shenyang Deng +2
Training strategies for modern deep neural networks (NNs) tend to induce a heavy-tailed (HT) empirical spectral density (ESD) in the layer weights. While previous efforts have show…