ImprovingDeepLearningOptimizationthrough ConstrainedParameterRegularization

Feb-8-2026, 01:59:18 GMT–Neural Information Processing Systems

Unlike the uniform application of asingle penalty,CPR enforces an upper bound on astatistical measure, suchas theL2-norm, ofindividual parameter matrices.

large language model, machine learning, natural language, (20 more...)

Neural Information Processing Systems

Feb-8-2026, 01:59:18 GMT

Conferences PDF

Country:
- Europe > Germany
  - Baden-Württemberg > Karlsruhe Region
    - Heidelberg (0.04)
    - Karlsruhe (0.04)
- North America
  - Canada > Ontario
    - Toronto (0.14)
  - United States (0.04)

Genre:
- Research Report > New Finding (0.67)

Technology:
- Information Technology > Artificial Intelligence
  - Machine Learning > Neural Networks
    - Deep Learning (0.69)
  - Natural Language > Large Language Model (0.69)
  - Representation & Reasoning (0.68)
  - Vision (0.68)

Duplicate Docs Excel Report

Title
Improving Deep Learning Optimization through Constrained Parameter Regularization

Similar Docs Excel Report more

Title	Similarity	Source
None found