Metraipolis › archive › dc_da8a27ba64bd

WASABI: Wasserstein Sampling via Bridge Interpolants — arXiv:2610.41958

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.41958
License (as archived): CC0-1.0
Archived: 2026-11-02T07:51:25.607Z
SHA-256 of the archived text: 897c7eaec21f5c9137b7ea163abb6ff674ee46a6c86670d6fba11ba36f33e0ec
Stable URL of this page: https://metraipolis.com/documento/dc_da8a27ba64bd

How to cite

WASABI: Wasserstein Sampling via Bridge Interpolants — arXiv:2610.41958 — arXiv (en), https://arxiv.org/abs/2610.41958 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_da8a27ba64bd (Metraipolis archive dc_da8a27ba64bd, sha256 897c7eaec21f5c9137b7ea163abb6ff674ee46a6c86670d6fba11ba36f33e0ec)

BibTeX

@misc{dcda8a27ba64bd,
  title = {WASABI: Wasserstein Sampling via Bridge Interpolants — arXiv:2610.41958},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.41958},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_da8a27ba64bd; sha256 897c7eaec21f5c9137b7ea163abb6ff674ee46a6c86670d6fba11ba36f33e0ec},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41958. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Sampling from Boltzmann distributions is a fundamental problem in physics, chemistry, and biology. We introduce WAsserstein SAmpling via Bridge Interpolants (WASABI), a method that trains a diffusion model to generate samples from a target Boltzmann distribution given access only to the energy function. WASABI minimizes the KL divergence between the distribution generated by its diffusion model and the Boltzmann target using Wasserstein gradient flows. This approach builds on a principle underlying classical sampling methods such as Langevin dynamics. However, applying Wasserstein gradient flows to diffusion models is challenging because the updates require the gradient of the model's own log density, which is generally intractable. We derive a practical training algorithm that avoids computing this gradient. Once trained, WASABI can generate new samples without further evaluations of the energy or its gradient. On alanine dipeptide, WASABI outperforms the best reported baselines using an estimated 3.3 times fewer energy evaluations. It also achieves competitive performance on interacting particle system benchmarks.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.41958 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-02T07:40:00.274Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).