How Does Gradient Descent Learn Features -- A Local Analysis for Regularized Two-Layer Neural Networks

Ähnliche Einträge

• Feature Learning in Linear-Width Two-Layer Networks: Two vs. One Step of Gradient Descent
  von: Moniri, Behrad, et al.
  Veröffentlicht: (2026)
• Generalization Guarantees of Gradient Descent for Multi-Layer Neural Networks
  von: Wang, Puyu, et al.
  Veröffentlicht: (2023)
• Stochastic Gradient Descent for Two-layer Neural Networks
  von: Cao, Dinghao, et al.
  Veröffentlicht: (2024)
• Convergence of Implicit Gradient Descent for Training Two-Layer Physics-Informed Neural Networks
  von: Xu, Xianliang, et al.
  Veröffentlicht: (2024)
• A Theory of Non-Linear Feature Learning with One Gradient Step in Two-Layer Neural Networks
  von: Moniri, Behrad, et al.
  Veröffentlicht: (2023)