A Random-Matrix Criterion for Initializing Gated Recurrent Neural Networks

Documenti analoghi

• Dynamical Learning in Deep Asymmetric Recurrent Neural Networks
  di: Badalotti, Davide, et al.
  Pubblicazione: (2025)
• Deep Linear Network Training Dynamics from Random Initialization: Data, Width, Depth, and Hyperparameter Transfer
  di: Bordelon, Blake, et al.
  Pubblicazione: (2025)
• Controlled Langevin Dynamics for Sampling of Feedforward Neural Networks Trained with Minibatches
  di: Zambon, Alessandro, et al.
  Pubblicazione: (2026)
• Small Singular Values Matter: A Random Matrix Analysis of Transformer Models
  di: Staats, Max, et al.
  Pubblicazione: (2024)
• Where You Place the Norm Matters: From Prejudiced to Neutral Initializations
  di: Francazi, Emanuele, et al.
  Pubblicazione: (2025)