No article text was recorded for this version.
theoretical-practical width gap
Version 2 of 2359 papers450 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Bounding the Width of Neural Networks via Coupled Initialization - A Worst Case Analysis
- Blog: Bounding the Width of Neural Networks via Coupled Initialization A Worst Case Analysis
- Blog: An Over-parameterized Exponential Regression
- Blog: Towards moderate overparameterization: global convergence guarantees for training shallow neural networks
- Blog: A Convergence Theory Towards Practical Over-parameterized Deep Neural Networks
- Blog: On the Convergence of Deep Networks with Sample Quadratic Overparameterization
- Blog: Quadratic Suffices for Over-parametrization via Matrix Chernoff Bound
- Blog: Solving Neural Min-Max Games: The Role of Architecture, Initialization & Dynamics
- Blog: Memorization and Optimization in Deep Neural Networks with Minimum Over-parameterization
- Blog: Global Convergence of Deep Networks with One Wide Layer Followed by Pyramidal Topology
- Blog: An Improved Analysis of Training Over-parameterized Deep Neural Networks
- Blog: Efficient NTK using Dimensionality Reduction
- Blog: Convolutional Neural Bandit: Provable Algorithm for Visual-aware Advertising
- Blog: Convergence of the Gradient Flow for Shallow ReLU Networks on Weakly Interacting Data
- Blog: An Improved Analysis of Training Over-parameterized Deep Neural Networks
- Blog: How Much Over-parameterization Is Sufficient to Learn Deep ReLU Networks?
- Blog: Gradient Descent Optimizes Infinite-Depth ReLU Implicit Networks with Linear Widths
- Blog: Data-driven learning of feedback maps for explicit robust predictive control: an approximation theoretic view
- Blog: Early Stopping of Untrained Convolutional Neural Networks
- Blog: Provable Accelerated Convergence of Nesterov's Momentum for Deep ReLU Neural Networks
- Blog: Overparameterization of deep ResNet: zero loss and mean-field analysis
- Blog: Convergence of Gradient Descent for Recurrent Neural Networks: A Nonasymptotic Analysis
- Blog: Solving Neural Min-Max Games: The Role of Architecture, Initialization & Dynamics
- Blog: On the Generalization Properties of Minimum-norm Solutions for Over-parameterized Neural Network Models
- Blog: Gradient Descent Provably Optimizes Over-parameterized Neural Networks
- Blog: Combinatorial Neural Bandits
- Blog: Training (Overparametrized) Neural Networks in Near-Linear Time
- Blog: Stochastic Gradient Descent Optimizes Over-parameterized Deep ReLU Networks
- Blog: On Learning Over-parameterized Neural Networks: A Functional Approximation Perspective
- Blog: Exponentially vanishing sub-optimal local minima in multilayer neural networks
- Blog: Convergence of End-to-End Training in Deep Unsupervised Contrasitive Learning
- Blog: On the Convergence of Encoder-only Shallow Transformers
- Blog: Neural Networks Learning and Memorization with (almost) no Over-Parameterization
- Blog: The Dynamics of Gradient Descent for Overparametrized Neural Networks
- Blog: Neural Active Learning Beyond Bandits
- Blog: On the Power of Small-size Graph Neural Networks for Linear Programming
- Blog: Dynamics of Deep Neural Networks and Neural Tangent Hierarchy
- Blog: No bad local minima: Data independent training error guarantees for multilayer neural networks
- Blog: Over-Parameterized Deep Neural Networks Have No Strict Local Minima For Any Continuous Activations
- Blog: Neural Contextual Bandits with Deep Representation and Shallow Exploration
- Blog: Global Convergence in Neural ODEs: Impact of Activation Functions
- Blog: On the optimization and generalization of overparameterized implicit neural networks
- Blog: No Spurious Local Minima in Deep Quadratic Networks
- Blog: Provably and Practically Efficient Neural Contextual Bandits
- Blog: Learning Lipschitz Functions by GD-trained Shallow Overparameterized ReLU Neural Networks
- Blog: Stochastic Gradient Descent for Two-layer Neural Networks
- Blog: Generalization and Stability of Interpolating Neural Networks with Minimal Width
- Blog: Expressivity of Neural Networks with Random Weights and Learned Biases.
- Blog: A Dynamical View on Optimization Algorithms of Overparameterized Neural Networks
- Blog: Bounding the Width of Neural Networks via Coupled Initialization A Worst Case Analysis
- Blog: On the Effective Number of Linear Regions in Shallow Univariate ReLU Networks: Convergence Guarantees and Implicit Bias
- Blog: NeuralBO: A black-box optimization algorithm using deep neural networks
- Blog: On Connected Sublevel Sets in Deep Learning
- Blog: Achieving Small Test Error in Mildly Overparameterized Neural Networks
- Blog: Refined Generalization Analysis of Gradient Descent for Over-parameterized Two-layer Neural Networks with Smooth Activations on Classification Problems
- Blog: On the Global Convergence of Gradient Descent for multi-layer ResNets in the mean-field regime
- Blog: Optimal Rates for Averaged Stochastic Gradient Descent under Neural Tangent Kernel Regime
- Blog: Local SGD Optimizes Overparameterized Neural Networks in Polynomial Time
- Blog: Nearly Minimal Over-Parametrization of Shallow Neural Networks
- Blog: No Spurious Local Minima: on the Optimization Landscapes of Wide and Deep Neural Networks
Showing 60 of 359 papers.
History & citation
Current · Permanent link
- Tightening the Width Bounds for Global Convergence merged into this direction (83% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_theoretical_practical_width_gap_ce521080,
title = {theoretical-practical width gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/theoretical-practical-width-gap/rdr_ce521080f4a54d2cab86ca21f73d0108},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}This exact version (version 2, frozen text):
@misc{lacuna_theoretical_practical_width_gap_ce521080_v2,
title = {theoretical-practical width gap},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/theoretical-practical-width-gap/rdr_ce521080f4a54d2cab86ca21f73d0108/v/2},
note = {Version 2, frozen 2026-10-04. Accessed 2026-11-20},
}