Imagen de Portada

Evaluating Singular Value Thresholds for DNN Weight Matrices based on Random Matrix Theory

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Nishikawa, Kohei, Shimizu, Koki, Hashiguchi, Hiroki
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

Exact Distribution of the Noncentral Complex Roy's Largest Root Statistic via Pieri's Formula
por: Shimizu, Koki, et al.
Publicado: (2025)

Perturbation Analysis of Singular Values in Concatenated Matrices
por: Shamrai, Maksym
Publicado: (2025)

Higher-Order Singular-Value Derivatives of Rectangular Real Matrices
por: Luo, Róisín, et al.
Publicado: (2025)

Small Singular Values Matter: A Random Matrix Analysis of Transformer Models
por: Staats, Max, et al.
Publicado: (2024)

Mining of Switching Sparse Networks for Missing Value Imputation in Multivariate Time Series
por: Obata, Kohei, et al.
Publicado: (2024)

A Generalized Singular Value Theory for Neural Networks
por: Brown, Brian Charles, et al.
Publicado: (2026)

A Theory of Diversity for Random Matrices with Applications to In-Context Learning of Schrödinger Equations
por: Cole, Frank, et al.
Publicado: (2026)

Neural Tangent Kernels and Fisher Information Matrices for Simple ReLU Networks with Random Hidden Weights
por: Takeuchi, Jun'ichi, et al.
Publicado: (2025)

A Random Matrix Theory of Masked Self-Supervised Regression
por: Zurich, Arie Wortsman, et al.
Publicado: (2026)

A Robust and Non-Iterative Tensor Decomposition Method with Automatic Thresholding
por: Hasegawa, Hiroki, et al.
Publicado: (2025)

Random Matrix Theory for Deep Learning: Beyond Eigenvalues of Linear Models
por: Liao, Zhenyu, et al.
Publicado: (2025)

Enhancing Accuracy in Deep Learning Using Random Matrix Theory
por: Berlyand, Leonid, et al.
Publicado: (2023)

Matrix-Valued Optimism is Matrix-Valued Augmentation: Additive Hybrid Designs for Constrained Optimization
por: Zhao, Jiayi
Publicado: (2026)

FLoRIST: Singular Value Thresholding for Efficient and Accurate Federated Fine-Tuning of Large Language Models
por: Ramesh, Hariharan, et al.
Publicado: (2025)

Beyond the Singular: Revealing the Value of Multiple Generations in Benchmark Evaluation
por: Zhang, Wenbo, et al.
Publicado: (2025)

EDoRA: Efficient Weight-Decomposed Low-Rank Adaptation via Singular Value Decomposition
por: Nasiri, Hamid, et al.
Publicado: (2025)

A Random Matrix Theory Perspective on the Learning Dynamics of Multi-head Latent Attention
por: Jha, Nandan Kumar, et al.
Publicado: (2025)

On Reward Transferability in Adversarial Inverse Reinforcement Learning: Insights from Random Matrix Theory
por: Zhang, Yangchun, et al.
Publicado: (2024)

Applications of Random Matrix Theory in Machine Learning and Brain Mapping
por: Lawrence, Katrina
Publicado: (2025)

SEMU: Singular Value Decomposition for Efficient Machine Unlearning
por: Sendera, Marcin, et al.
Publicado: (2025)

Quantifying the Uncertainty of Foundation Models with Singular Value Ensembles
por: Turkoglu, Mehmet Ozgur, et al.
Publicado: (2026)

Optimal Estimation of Shared Singular Subspaces across Multiple Noisy Matrices
por: Ma, Zhengchi, et al.
Publicado: (2024)

Analysing Multi-Task Regression via Random Matrix Theory with Application to Time Series Forecasting
por: Ilbert, Romain, et al.
Publicado: (2024)

Beyond Accuracy: A Unified Random Matrix Theory Diagnostic Framework for Crash Classification Models
por: Shihab, Ibne Farabi, et al.
Publicado: (2026)

Element-wise Modulation of Random Matrices for Efficient Neural Layers
por: Szorc, Maksymilian
Publicado: (2025)

Scalable Unsupervised Segmentation via Random Fourier Feature-based Gaussian Process
por: Saito, Issei, et al.
Publicado: (2025)

Generalization in Representation Models via Random Matrix Theory: Application to Recurrent Networks
por: Moakher, Yessin, et al.
Publicado: (2025)

Skip2-LoRA: A Lightweight On-device DNN Fine-tuning Method for Low-cost Edge Devices
por: Matsutani, Hiroki, et al.
Publicado: (2024)

Stochastic Rounding Increases Small Singular Values
por: Ma, Linkai, et al.
Publicado: (2026)

SORSA: Singular Values and Orthonormal Regularized Singular Vectors Adaptation of Large Language Models
por: Cao, Yang, et al.
Publicado: (2024)

PiSSA: Principal Singular Values and Singular Vectors Adaptation of Large Language Models
por: Meng, Fanxu, et al.
Publicado: (2024)

Manifolds, Random Matrices and Spectral Gaps: The geometric phases of generative diffusion
por: Ventura, Enrico, et al.
Publicado: (2024)

Learning with Density Matrices and Random Features
por: González, Fabio A., et al.
Publicado: (2021)

Random Matrix Theory for Stochastic Gradient Descent
por: Park, Chanju, et al.
Publicado: (2024)

Spectral Geometry for Deep Learning: Compression and Hallucination Detection via Random Matrix Theory
por: Ettori, Davide
Publicado: (2026)

A Random Matrix Theory Perspective on the Spectrum of Learned Features and Asymptotic Generalization Capabilities
por: Dandi, Yatin, et al.
Publicado: (2024)

Random Matrix Theory of Early-Stopped Gradient Flow: A Transient BBP Scenario
por: Coeurdoux, Florentin, et al.
Publicado: (2026)

Dynamic Multi-Network Mining of Tensor Time Series
por: Obata, Kohei, et al.
Publicado: (2024)

Maximizing the Potential of Synthetic Data: Insights from Random Matrix Theory
por: Firdoussi, Aymane El, et al.
Publicado: (2024)

Two-Valued Symmetric Circulant Matrices: Applications in Deep Learning
por: Amathi, Jayakrishna, et al.
Publicado: (2026)