Product
Help
Pricing
About
Jobs
Log in
Get started
Keyword
Add to collection
infinite-width limit
Description:
An asymptotic regime in neural network theory where the number of units in each layer grows without bound.
Primary topic:
Neural Networks and Applications (32%)
Related topics:
Neural Networks and Applications (32%),
Gaussian Processes and Bayesian Inference (20%),
Stochastic Gradient Optimization Techniques (20%)
OpenAlex ID:
keywords/infinite-width-limit
When and why PINNs fail to train: A neural tangent kernel perspective
2021
·
Sifan Wang
, Xinling Yu
, et al.
·
Journal of Computational Physics
·
2,439
PDF
Neural Tangent Kernel: Convergence and Generalization in Neural Networks
2018
·
Jacot, Arthur
, Gabriel, Franck
, et al.
·
neural information processing systems
·
2,073
On Exact Computation with an Infinitely Wide Neural Net
2019
·
Arora, Sanjeev
, Du, Simon S.
, et al.
·
neural information processing systems
·
1,024
Wide neural networks of any depth evolve as linear models under gradient descent *
2020
·
Jaehoon Lee
, Lechao Xiao
, et al.
·
Journal of Statistical Mechanics Theory and Experiment
·
775
PDF
Explaining neural scaling laws
2024
·
Yasaman Bahri
, Ethan Dyer
, et al.
·
Proceedings of the National Academy of Sciences
·
665
Spectral bias and task-model alignment explain generalization in kernel regression and infinitely wide neural networks
2021
·
Abdulkadir Canatar
, Blake Bordelon
, et al.
·
Nature Communications
·
323
PDF
Linearized two-layers neural networks in high dimension
2021
·
Behrooz Ghorbani
, Song Mei
, et al.
·
The Annals of Statistics
·
313
PDF
Deep Convolutional Networks as shallow Gaussian Processes
2018
·
Garriga-Alonso, Adrià
, Aitchison, Laurence
, et al.
·
arXiv (Cornell University)
·
248
PDF
Bayesian Deep Ensembles via the Neural Tangent Kernel
2020
·
He, Bobby
, Lakshminarayanan, Balaji
, et al.
·
neural information processing systems
·
133
The Spectrum of the Fisher Information Matrix of a Single-Hidden-Layer Neural Network
2018
·
Jeffrey Pennington
, Pratik Worah
·
neural information processing systems
·
102
Phase diagram for two-layer ReLU neural networks at infinite-width limit
2020
·
Luo, Tao
, Xu, Zhi-Qin John
, et al.
·
arXiv (Cornell University)
·
101
PDF
A statistical mechanics framework for Bayesian deep neural networks beyond the infinite-width limit
2023
·
R. Pacelli
, S. Ariosto
, et al.
·
Nature Machine Intelligence
·
95
PDF
Graph Neural Networks are Inherently Good Generalizers: Insights by Bridging GNNs and MLPs
2022
·
Yang, Chenxiao
, Wu, Qitian
, et al.
·
arXiv (Cornell University)
·
93
PDF
The Recurrent Neural Tangent Kernel
2020
·
Alemohammad, Sina
, Wang, Zichao
, et al.
·
arXiv (Cornell University)
·
86
PDF
Depthwise Hyperparameter Transfer in Residual Networks: Dynamics and Scaling Limit
2024
·
Bordelon, Blake
, Noci, Lorenzo
, et al.
·
International Conference on Learning Representations
·
82
PDF
Random neural networks in the infinite width limit as Gaussian processes
2023
·
Boris Hanin
·
The Annals of Applied Probability
·
82
PDF
Tensor Programs IIb: Architectural Universality of Neural Tangent Kernel Training Dynamics
2021
·
Yang, Greg
, Littwin, Etai
·
arXiv (Cornell University)
·
76
PDF
Jellium surface energy beyond the local-density approximation: Self-consistent-field calculations
2001
·
J. Pitarke
, A. Eguiluz
·
Physical review. B, Condensed matter
·
73
PDF
Predicting the outputs of finite deep neural networks trained with noisy gradients
2021
·
Gadi Naveh
, Oded Ben David
, et al.
·
Physical review. E
·
68
PDF
Infinite Limits of Multi-head Transformer Dynamics
2024
·
Blake Bordelon
, Hamza Chaudhry
, et al.
·
neural information processing systems
·
65
Non-Gaussian processes and neural networks at finite widths
2019
·
Yaida, Sho
·
arXiv (Cornell University)
·
65
PDF
Non-perturbative renormalization for the neural network-QFT correspondence
2022
·
H Erbin
, V Lahoche
, et al.
·
Machine Learning Science and Technology
·
64
PDF
Towards Understanding Hierarchical Learning: Benefits of Neural Representations
2020
·
Chen, Minshuo
, Bai, Yu
, et al.
·
neural information processing systems
·
62
Why Do Deep Residual Networks Generalize Better than Deep Feedforward Networks? -- A Neural Tangent Kernel Perspective
2020
·
Huang, Kaixuan
, Wang, Yuqing
, et al.
·
neural information processing systems
·
56
Symmetry-via-Duality: Invariant Neural Network Densities from Parameter-Space Correlators
2023
·
Anindita Maiti
, Keegan Stoner
, et al.
·
WORLD SCIENTIFIC (EUROPE) eBooks
·
49
PDF
Show more works
View as search
Works count:
208
Citation count:
10,860
Year
Topic
Type