Imagen de Portada

Asymptotic Analysis of Two-Layer Neural Networks after One Gradient Step under Gaussian Mixtures Data with Structure

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Demir, Samet, Dogan, Zafer
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

How Data Mixing Shapes In-Context Learning: Asymptotic Equivalence for Transformers with MLPs
por: Demir, Samet, et al.
Publicado: (2025)

Learning Beyond the Gaussian Data: Learning Dynamics of Neural Networks on an Expressive and Cumulant-Controllable Data Model
por: Ure, Onat, et al.
Publicado: (2026)

Asymptotic Study of In-context Learning with Random Transformers through Equivalent Models
por: Demir, Samet, et al.
Publicado: (2025)

Optimal Attention Temperature Improves the Robustness of In-Context Learning under Distribution Shift in High Dimensions
por: Demir, Samet, et al.
Publicado: (2025)

Input-Label Correlation Governs a Linear-to-Nonlinear Transition in Random Features under Spiked Covariance
por: Demir, Samet, et al.
Publicado: (2024)

Implicitly Normalized Online PCA: A Regularized Algorithm with Exact High-Dimensional Dynamics
por: Demir, Samet, et al.
Publicado: (2025)

Learning Rate Should Scale Inversely with High-Order Data Moments in High-Dimensional Online Independent Component Analysis
por: Gultekin, M. Oguzhan, et al.
Publicado: (2025)

Benefits of Online Tilted Empirical Risk Minimization: A Case Study of Outlier Detection and Robust Regression
por: Yildirim, Yigit E., et al.
Publicado: (2025)

Learnability and Competition in High-Dimensional Multi-Component ICA
por: Genc, Eser Ilke, et al.
Publicado: (2026)

A Theory of Non-Linear Feature Learning with One Gradient Step in Two-Layer Neural Networks
por: Moniri, Behrad, et al.
Publicado: (2023)

Exploring the Precise Dynamics of Single-Layer GAN Models: Leveraging Multi-Feature Discriminators for High-Dimensional Subspace Learning
por: Bond, Andrew, et al.
Publicado: (2024)

Feature Learning in Linear-Width Two-Layer Networks: Two vs. One Step of Gradient Descent
por: Moniri, Behrad, et al.
Publicado: (2026)

How Two-Layer Neural Networks Learn, One (Giant) Step at a Time
por: Dandi, Yatin, et al.
Publicado: (2023)

Convergence Analysis of Two-Layer Neural Networks under Gaussian Input Masking
por: Kolomvaki, Afroditi, et al.
Publicado: (2026)

How Does Gradient Descent Learn Features -- A Local Analysis for Regularized Two-Layer Neural Networks
por: Zhou, Mo, et al.
Publicado: (2024)

Step by Step: Adaptive Gradient Descent for Training L-Lipschitz Neural Networks
por: Sung, Kyle, et al.
Publicado: (2025)

Latent Chain-of-Thought Improves Structured-Data Transformers
por: Dudley, Carson, et al.
Publicado: (2026)

One-Pass Sparsified Gaussian Mixtures
por: Kightley, Eric, et al.
Publicado: (2019)

Balancing Learning Rates Across Layers: Exact Two-Step Dynamics and Optimal Scaling in Linear Neural Networks
por: Pang, Tianyu, et al.
Publicado: (2026)

Convergence of Stochastic Gradient Methods for Wide Two-Layer Physics-Informed Neural Networks
por: Jin, Bangti, et al.
Publicado: (2025)

Convergence of Implicit Gradient Descent for Training Two-Layer Physics-Informed Neural Networks
por: Xu, Xianliang, et al.
Publicado: (2024)

A Gap Between the Gaussian RKHS and Neural Networks: An Infinite-Center Asymptotic Analysis
por: Kumar, Akash, et al.
Publicado: (2025)

uGMM-NN: Univariate Gaussian Mixture Model Neural Network
por: Ali, Zakeria Sharif
Publicado: (2025)

Generalization Guarantees of Gradient Descent for Multi-Layer Neural Networks
por: Wang, Puyu, et al.
Publicado: (2023)

Provable Unlearning with Gradient Ascent on Two-Layer ReLU Neural Networks
por: Melamed, Odelia, et al.
Publicado: (2025)

Stochastic Gradient Descent for Two-layer Neural Networks
por: Cao, Dinghao, et al.
Publicado: (2024)

A Unified Framework for Data-Free One-Step Sampling via Wasserstein Gradient Flows
por: Wang, Chenguang, et al.
Publicado: (2026)

Asymptotic Smoothing of the Lipschitz Loss Landscape in Overparameterized One-Hidden-Layer ReLU Networks
por: Baturin, Saveliy
Publicado: (2026)

Diffusion Model Conditioning on Gaussian Mixture Model and Negative Gaussian Mixture Gradient
por: Lu, Weiguo, et al.
Publicado: (2024)

Analyzing Neural Scaling Laws in Two-Layer Networks with Power-Law Data Spectra
por: Worschech, Roman, et al.
Publicado: (2024)

Global Convergence of Gradient EM for Over-Parameterized Gaussian Mixtures
por: Zhou, Mo, et al.
Publicado: (2025)

Rethinking Benign Overfitting in Two-Layer Neural Networks
por: Xu, Ruichen, et al.
Publicado: (2025)

Effective Capacitance Modeling Using Graph Neural Networks
por: Dogan, Eren, et al.
Publicado: (2025)

Analytic Convolutional Layer: A Step to Analytic Neural Network
por: Cui, Jingmao, et al.
Publicado: (2024)

A Lightweight and Gradient-Stable Neural Layer
por: Yu, Yueyao, et al.
Publicado: (2021)

Finite Neural Networks as Mixtures of Gaussian Processes: From Provable Error Bounds to Prior Selection
por: Adams, Steven, et al.
Publicado: (2024)

The Double Descent Behavior in Two Layer Neural Network for Binary Classification
por: Abeykoon, Chathurika S, et al.
Publicado: (2025)

AdaBet: Gradient-free Layer Selection for Efficient Training of Deep Neural Networks
por: Tenison, Irene, et al.
Publicado: (2025)

Structured Diffusion Models with Mixture of Gaussians as Prior Distribution
por: Jia, Nanshan, et al.
Publicado: (2024)

Training of Neural Networks with Uncertain Data: A Mixture of Experts Approach
por: Luttner, Lucas
Publicado: (2023)