Inversion dynamics of class manifolds in deep learning reveals tradeoffs underlying generalisation

Documenti analoghi

• Feature learning in finite-width Bayesian deep linear networks with multiple outputs and convolutional layers
  di: Bassetti, Federico, et al.
  Pubblicazione: (2024)
• Differential privacy for medical deep learning: methods, tradeoffs, and deployment implications
  di: Mohammadi, Marziyeh, et al.
  Pubblicazione: (2025)
• Carré du champ flow matching: better quality-generalisation tradeoff in generative models
  di: Bamberger, Jacob, et al.
  Pubblicazione: (2025)
• Microscopic and collective signatures of feature learning in neural networks
  di: Corti, Andrea, et al.
  Pubblicazione: (2025)
• Should We Always Train Models on Fine-Grained Classes?
  di: Pirovano, Davide, et al.
  Pubblicazione: (2025)