EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Mean-field limit from general mixtures of experts to quantum neural networks

Anderson Melchor Hernandez, Davide Pastorello, Giacomo De Palma

Letters in Mathematical Physics · 27.03.2026 · 116 (2)

DOI: 10.1007/s11005-026-02065-9 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Letters in Mathematical Physics, online erschienen am 27.03.2026. Verfasst von Anderson Melchor Hernandez und 2 weiteren Autor:innen. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: PNRR Prin 2022, PNRR.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In this work, we study the asymptotic behavior of mixture of experts (MoE) trained via gradient flow on supervised learning problems. Our main result establishes the propagation of chaos for a MoE as the number of experts diverges. We demonstrate that the corresponding empirical measure of their parameters is close to a probability measure that solves a nonlinear continuity equation, and we provide an explicit convergence rate that depends solely on the number of experts and on the dimensionality of the parameter space. We apply our results to a MoE generated by a quantum neural network.

Förderung

  • PNRR Prin 2022
  • PNRR

Zitieren

APA

Hernandez, A. M., Pastorello, D., & Palma, G. D. (2026). Mean-field limit from general mixtures of experts to quantum neural networks. Letters in Mathematical Physics, 116(2). https://doi.org/10.1007/s11005-026-02065-9

Vancouver

Hernandez AM, Pastorello D, Palma GD. Mean-field limit from general mixtures of experts to quantum neural networks. Letters in Mathematical Physics. 2026;116(2). doi:10.1007/s11005-026-02065-9

BibTeX
@article{Hernandez2026,
  author = {Hernandez, Anderson Melchor and Pastorello, Davide and Palma, Giacomo De},
  title = {Mean-field limit from general mixtures of experts to quantum neural networks},
  journal = {Letters in Mathematical Physics},
  year = {2026},
  volume = {116},
  number = {2},
  doi = {10.1007/s11005-026-02065-9}
}
Kostenlos starten →