What does guidance do? A fine-grained analysis in a simple setting

Documenti analoghi

• For Better or For Worse? Learning Minimum Variance Features With Label Augmentation
  di: Chidambaram, Muthu, et al.
  Pubblicazione: (2024)
• Critical windows: non-asymptotic theory for feature emergence in diffusion models
  di: Li, Marvin, et al.
  Pubblicazione: (2024)
• Provably Learning Diverse Features in Multi-View Data with Midpoint Mixup
  di: Chidambaram, Muthu, et al.
  Pubblicazione: (2022)
• High-accuracy and dimension-free sampling with diffusions
  di: Gatmiry, Khashayar, et al.
  Pubblicazione: (2026)
• When does Metropolized Hamiltonian Monte Carlo provably outperform Metropolis-adjusted Langevin algorithm?
  di: Chen, Yuansi, et al.
  Pubblicazione: (2023)