Global Convergence of Gradient Descent for Deep Linear Residual Networks

Lei Wu, Qingcan Wang, Chao Ma

Feb-11-2026, 13:16:47 GMT–Neural Information Processing Systems

It is motivated by avoiding stable manifolds of saddle points.

artificial intelligence, initialization, machine learning, (18 more...)

Neural Information Processing Systems

Feb-11-2026, 13:16:47 GMT

Conferences PDF

Country:
- North America
  - Canada (0.04)
  - United States > New Jersey
    - Mercer County > Princeton (0.04)

Technology:
- Information Technology > Artificial Intelligence > Machine Learning
  - Neural Networks > Deep Learning (0.47)
  - Statistical Learning > Gradient Descent (0.59)

Duplicate Docs Excel Report

Title
Global Convergence of Gradient Descent for Deep Linear Residual Networks

Similar Docs Excel Report more

Title	Similarity	Source
None found