New study examines point interpolation and distributional universality in flow-based generative models
A new arXiv paper studies the approximation of probability distributions in flow-based generative models, motivated by the expressive power of neural ordinary differential equations.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-04 05:00 · arXiv stat.ML
Point interpolation and distributional universality in flow-based generative models