Imagen de Portada

The Spectral Dimension of NTKs is Constant: A Theory of Implicit Regularization, Finite-Width Stability, and Scalable Estimation

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autor principal:	Shukla, Praveen Anilkumar
Formato:	Preprint
Publicado:	2025
Materias:	Machine Learning
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Implicit Regularization in Perturbed Deep Matrix Factorization: Spectral Conditions and Stability
por: Wang, Jingzhe, et al.
Publicado: (2026)

Some Theoretical Results on Layerwise Effective Dimension Oscillations in Finite Width ReLU Networks
por: Makwana, Darshan
Publicado: (2025)

Estimating Implicit Regularization in Deep Learning
por: Rudoler, Joseph H., et al.
Publicado: (2026)

Local Hessian Spectral Filtering for Robust Intrinsic Dimension Estimation
por: Osada, Genki
Publicado: (2026)

Efficient Techniques for Data Reconstruction, with Finite-Width Recovery Guarantees
por: Tansley, Edward, et al.
Publicado: (2026)

Spectral Condition for $μ$P under Width-Depth Scaling
por: Zheng, Chenyu, et al.
Publicado: (2026)

The Implicit Bias of Logit Regularization
por: Beck, Alon, et al.
Publicado: (2026)

Channel Estimation by Infinite Width Convolutional Networks
por: Mallik, Mohammed, et al.
Publicado: (2025)

Scalable Implicit Graphon Learning
por: Azizpour, Ali, et al.
Publicado: (2024)

Risk Comparisons in Linear Regression: Implicit Regularization Dominates Explicit Regularization
por: Wu, Jingfeng, et al.
Publicado: (2025)

Implicit Regularization via Spectral Neural Networks and Non-linear Matrix Sensing
por: Chu, Hong T. M., et al.
Publicado: (2024)

Network Degeneracy as an Indicator of Training Performance: Comparing Finite and Infinite Width Angle Predictions
por: Jakub, Cameron, et al.
Publicado: (2023)

Spectral Thresholds for Identifiability and Stability:Finite-Sample Phase Transitions in High-Dimensional Learning
por: Huang, William Hao-Cheng
Publicado: (2025)

Learning Continually by Spectral Regularization
por: Lewandowski, Alex, et al.
Publicado: (2024)

Spectral Regularization for Diffusion Models
por: Chandran, Satish, et al.
Publicado: (2026)

Finite-Width Neural Tangent Kernels from Feynman Diagrams
por: Guillen, Max, et al.
Publicado: (2025)

The Median is Easier than it Looks: Approximation with a Constant-Depth, Linear-Width ReLU Network
por: Dutta, Abhigyan, et al.
Publicado: (2026)

Scaled Supervision is an Implicit Lipschitz Regularizer
por: Ouyang, Zhongyu, et al.
Publicado: (2025)

Implicit Regularization of the Deep Inverse Prior Trained with Inertia
por: Buskulic, Nathan, et al.
Publicado: (2025)

Networks with Finite VC Dimension: Pro and Contra
por: Kurkova, Vera, et al.
Publicado: (2025)

A Unified Approach to Controlling Implicit Regularization via Mirror Descent
por: Sun, Haoyuan, et al.
Publicado: (2023)

From Sublinear to Linear: Local Convergence in Finite-Width Networks via Locally Polyak-Lojasiewicz Regions
por: Aich, Agnideep, et al.
Publicado: (2025)

On the Saturation Effects of Spectral Algorithms in Large Dimensions
por: Lu, Weihao, et al.
Publicado: (2025)

Latent Spectral Regularization for Continual Learning
por: Frascaroli, Emanuele, et al.
Publicado: (2023)

Piecewise Constant Spectral Graph Neural Network
por: Martirosyan, Vahan, et al.
Publicado: (2025)

Implicit Regularization in Feedback Alignment Learning Mechanisms for Neural Networks
por: Robertson, Zachary, et al.
Publicado: (2023)

Implicit Regularization of Sharpness-Aware Minimization for Scale-Invariant Problems
por: Li, Bingcong, et al.
Publicado: (2024)

Improving Implicit Regularization of SGD with Preconditioning for Least Square Problems
por: Su, Junwei, et al.
Publicado: (2024)

Implicit Regularization Towards Rank Minimization in ReLU Networks
por: Timor, Nadav, et al.
Publicado: (2022)

Variational Deep Learning via Implicit Regularization
por: Wenger, Jonathan, et al.
Publicado: (2025)

Implicit Regularization Paths of Weighted Neural Representations
por: Du, Jin-Hong, et al.
Publicado: (2024)

Improving Generalization and Convergence by Enhancing Implicit Regularization
por: Wang, Mingze, et al.
Publicado: (2024)

Robust Implicit Regularization via Weight Normalization
por: Chou, Hung-Hsu, et al.
Publicado: (2023)

Entropy-based Training Methods for Scalable Neural Implicit Sampler
por: Luo, Weijian, et al.
Publicado: (2023)

Towards Scalable Topological Regularizers
por: Wong, Hiu-Tung, et al.
Publicado: (2025)

Non-Convex Optimization with Spectral Radius Regularization
por: Sandler, Adam, et al.
Publicado: (2021)

Mirror, Mirror of the Flow: How Does Regularization Shape Implicit Bias?
por: Jacobs, Tom, et al.
Publicado: (2025)

Connectivity Shapes Implicit Regularization in Matrix Factorization Models for Matrix Completion
por: Bai, Zhiwei, et al.
Publicado: (2024)

Geometric Limits of Knowledge Distillation: A Minimum-Width Theorem via Superposition Theory
por: Sarkar, Nilesh, et al.
Publicado: (2026)

A Novel Approach for Intrinsic Dimension Estimation
por: Özçoban, Kadir, et al.
Publicado: (2025)