:: Library Catalog

Imagen de Portada

Guardado en:

Detalles Bibliográficos
Autor principal:	Xu, Yongzhong
Formato:	Preprint
Publicado:	2026
Materias:	Machine Learning
Acceso en línea:	https://arxiv.org/abs/2604.07380
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Spectral Edge Dynamics Reveal Functional Modes of Learning
por: Xu, Yongzhong
Publicado: (2026)

The Geometry of Multi-Task Grokking: Transverse Instability, Superposition, and Weight Decay Phase Structure
por: Xu, Yongzhong
Publicado: (2026)

Spectral Edge Dynamics of Training Trajectories: Signal--Noise Geometry Across Scales
por: Xu, Yongzhong
Publicado: (2026)

Spectral Edge Dynamics: An Analytical-Empirical Study of Phase Transitions in Neural Network Training
por: Xu, Yongzhong
Publicado: (2026)

Feature Repulsion and Spectral Lock-in: An Empirical Study of Two-Layer Network Grokking
por: Xu, Yongzhong
Publicado: (2026)

Gradient-Direction Sensitivity Reveals Linear-Centroid Coupling Hidden by Optimizer Trajectories
por: Xu, Yongzhong
Publicado: (2026)

Spectral Probe-Circuits: A Three-Step Recipe for Identifying Attention-Head Circuits in Pretrained Transformers
por: Xu, Yongzhong
Publicado: (2026)

The Spectral Lifecycle of Transformer Training: Transient Compression Waves, Persistent Spectral Gradients, and the Q/K--V Asymmetry
por: Liu, Yi
Publicado: (2026)

Global Low-Rank, Local Full-Rank: The Holographic Encoding of Learned Algorithms
por: Xu, Yongzhong
Publicado: (2026)

Low-Dimensional Execution Manifolds in Transformer Learning Dynamics: Evidence from Modular Arithmetic Tasks
por: Xu, Yongzhong
Publicado: (2026)

Early-Warning Signals of Grokking via Loss-Landscape Geometry
por: Xu, Yongzhong
Publicado: (2026)

Low-Dimensional and Transversely Curved Optimization Dynamics in Grokking
por: Xu, Yongzhong
Publicado: (2026)

When Do Attention Circuits Form? Developmental Trajectories of Capability and Attention-Sink Emergence Across Three 1B-ClassArchitectures
por: Xu, Yongzhong
Publicado: (2026)

Optimizer-Induced Low-Dimensional Drift and Transverse Dynamics in Transformer Training
por: Xu, Yongzhong
Publicado: (2026)

On the Overlooked Pitfalls of Weight Decay and How to Mitigate Them: A Gradient-Norm Perspective
por: Xie, Zeke, et al.
Publicado: (2020)

The Rank and Gradient Lost in Non-stationarity: Sample Weight Decay for Mitigating Plasticity Loss in Reinforcement Learning
por: Wu, Zihao, et al.
Publicado: (2026)

Theoretical Analysis of Sparse Optimization with Reparameterization, Weight Decay, and Adaptive Learning Rate
por: Xu, Huangyu, et al.
Publicado: (2026)

Correction of Decoupled Weight Decay
por: Chou, Jason Chuan-Chih
Publicado: (2025)

From $\log π$ to $π$: Taming Divergence in Soft Clipping via Bilateral Decoupled Decay of Probability Gradient Weight
por: Fu, Xiaoliang, et al.
Publicado: (2026)

Cautious Weight Decay
por: Chen, Lizhang, et al.
Publicado: (2025)

Learning to Forget: Continual Learning with Adaptive Weight Decay
por: Ramesh, Aditya A., et al.
Publicado: (2026)

Why Do We Need Weight Decay in Modern Deep Learning?
por: D'Angelo, Francesco, et al.
Publicado: (2023)

Uncertainty Weighted Gradients for Model Calibration
por: Lin, Jinxu, et al.
Publicado: (2025)

Predicting LLM Compression Degradation from Spectral Statistics
por: Xu, Mingxue
Publicado: (2026)

Deep Learning meets Nonparametric Regression: Are Weight-Decayed DNNs Locally Adaptive?
por: Zhang, Kaiqi, et al.
Publicado: (2022)

Rotational Equilibrium: How Weight Decay Balances Learning Across Neural Networks
por: Kosson, Atli, et al.
Publicado: (2023)

Solving Inverse Problems with Deep Linear Neural Networks: Global Convergence Guarantees for Gradient Descent with Weight Decay
por: Laus, Hannah, et al.
Publicado: (2025)

EntroLLM: Entropy Encoded Weight Compression for Efficient Large Language Model Inference on Edge Devices
por: Sanyal, Arnab, et al.
Publicado: (2025)

Weighted Temporal Decay Loss for Learning Wearable PPG Data with Sparse Clinical Labels
por: Chung, Yunsung, et al.
Publicado: (2026)

A Self-Attentive Meta-Optimizer with Group-Adaptive Learning Rates and Weight Decay
por: Zhao, JiangBo, et al.
Publicado: (2026)

Does Weight Decay Enhance Training Stability?
por: Saether, Marius, et al.
Publicado: (2026)

Task Weighting through Gradient Projection for Multitask Learning
por: Bohn, Christian, et al.
Publicado: (2024)

Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning
por: Shen, Junhao, et al.
Publicado: (2026)

DP-AdamW: Investigating Decoupled Weight Decay and Bias Correction in Private Deep Learning
por: Chooi, Jay, et al.
Publicado: (2025)

Weight Decay may matter more than muP for Learning Rate Transfer in Practice
por: Kosson, Atli, et al.
Publicado: (2025)

AlphaDecay: Module-wise Weight Decay for Heavy-Tailed Balancing in LLMs
por: He, Di, et al.
Publicado: (2025)

Towards Understanding Neural Collapse: The Effects of Batch Normalization and Weight Decay
por: Pan, Leyan, et al.
Publicado: (2023)

SGD and Weight Decay Secretly Minimize the Rank of Your Neural Network
por: Galanti, Tomer, et al.
Publicado: (2022)

Approaching Deep Learning through the Spectral Dynamics of Weights
por: Yunis, David, et al.
Publicado: (2024)

Neural Weight Compression for Language Models
por: Ryu, Jegwang, et al.
Publicado: (2025)