Metraipolis › archive › dc_13eba2074eda

Fixed-point acceleration on Neural Wasserstein manifolds — arXiv:2611.09723

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.09723
License (as archived)
CC0-1.0
Archived
2026-11-09T22:41:40.713Z
SHA-256 of the archived text
6bbf3a584f522a0b5902435968902cf62a434a3e052047e9aa5a0bdb5f972969
Stable URL of this page
https://metraipolis.com/documento/dc_13eba2074eda

How to cite

Fixed-point acceleration on Neural Wasserstein manifolds — arXiv:2611.09723 — arXiv (en), https://arxiv.org/abs/2611.09723 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_13eba2074eda (Metraipolis archive dc_13eba2074eda, sha256 6bbf3a584f522a0b5902435968902cf62a434a3e052047e9aa5a0bdb5f972969)
BibTeX
@misc{dc13eba2074eda,
  title = {Fixed-point acceleration on Neural Wasserstein manifolds — arXiv:2611.09723},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.09723},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_13eba2074eda; sha256 6bbf3a584f522a0b5902435968902cf62a434a3e052047e9aa5a0bdb5f972969},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.09723. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We propose a fixed-point acceleration method for optimization over probability measures, applicable to a variety of machine learning problems. We represent probability measures as pushforwards of a reference distribution through a neural network map and equip the parameter space with the induced pullback Wasserstein metric. This geometric structure allows natural gradient descent to be interpreted as a fixed-point iteration on a Riemannian manifold. We accelerate this iteration using Riemannian Anderson mixing method, which exploits historical residuals to capture approximate second-order information without forming Hessians. In the present parametric setting, the manifold is represented by Euclidean coordinates endowed with a point-dependent metric, simplifying retraction and vector transport. We evaluate the method on multimodal sampling, PDE-based inverse problems, aggregation dynamics, Bayesian logistic regression, and generative modeling. Across the tested tasks, Anderson mixing reduces the number of natural-gradient iterations and achieves competitive performance relative to SGD and Adam in terms of both loss-gradient evaluations and wall-clock time. These results suggest that fixed-point acceleration is an efficient tool for improving the computational efficiency of training on parametric Wasserstein methods.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Optimization and Control (math.OC).
Pagina del lavoro: https://arxiv.org/abs/2611.09723 — 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 2026-11-09T22:33:47.163Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).