Discrete distributions are learnable from metastable samples
Abstract Physically motivated stochastic dynamics are widely used to sample from high-dimensional distributions. However, such samplers often get trapped in metastable states, approximately sampling from a distribution that differs significantly from the desired stationary state. We rigorously show...
Zapisane w:
| Główni autorzy: | , , , |
|---|---|
| Format: | Artigo |
| Język: | Inglês |
| Wydane: |
Nature Portfolio
2026-07-01
|
| Seria: | Nature Communications |
| Dostęp online: | https://doi.org/10.1038/s41467-026-75439-1 |
| Etykiety: |
Nie ma etykietki, Dołącz pierwszą etykiete!
|