Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 57 min ago
PreprintarXiv licencearXiv

Uniform-in-Time and Uniform-in-Temperature Particle Approximations of Coulomb MMD Flows

Abstract

We study particle approximations of maximum mean discrepancy (MMD) flows with the Coulomb kernel on the torus 𝕋d, d≥2, driven by thermal noise of strength τ≥0. For smooth, strictly positive initial and target densities and i.i.d. initial positions, we prove that the expected Wasserstein-1 distance between the empirical measure and the mean-field solution is at most of order N−1/d for d≥3 and √((log N)/N) for d=2, uniformly in time and in the temperature τ≥0, including the deterministic case τ=0. At each fixed τ>0, we also prove propagation of chaos in total variation for any k-fold marginal distributions of particles, uniformly in time, with a bound of order √(k/τ) times the empirical convergence rate. Our proof uses the modulated free-energy method of Bresch, Jabin, and Wang to reduce the particle convergence estimate to a relaxation estimate that controls the time integral of the mean-field velocity gradient uniformly in τ≥0. Near zero temperature, where the smoothing effect of the noise degenerates, we obtain this estimate by recentering the dynamics at thermal equilibrium and using the damping generated by the Coulomb interaction.

Related

  1. From one particle to N particles: uniform transport-information and log-Sobolev inequalities for Coulomb gases on the torus
    A Guillin, W Liu, B Nectoux +1
    HALPreprintCC BY-NC-ND 4.0Full textAnalysis & PDEsMathematical physicsHALarXiv
  2. Uniform logarithmic Sobolev inequalities for the 2D Coulomb gas at the diffusive temperature scale
    Matthew Rosenzweig
    arXivPreprintarXiv licenceFull textAnalysis & PDEsMathematical physicsarXiv
  3. Uniform convergence of canonical discrete uniformization on equilateral approximations
    Aijin Lin, Chunlei Liu
    arXivPreprintCC BY 4.0Full textGeometry & topologyarXiv
  4. Product approximations for uniform spanning trees and determinantal processes
    András Mészáros, Yuval Peled
    arXivPreprintarXiv licenceFull textCombinatorics & discrete mathematicsProbabilityarXiv
  5. Weak solutions to granular medium equations via stochastic particle approximations
    Łukasz Wodnicki
    arXivPreprintCC BY 4.0Full textProbabilityarXiv
BibTeX
@misc{lee2026uniformintime,
  title = {{Uniform-in-Time and Uniform-in-Temperature Particle Approximations of Coulomb MMD Flows}},
  author = {Junho Lee and Promit Ghosal},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.15950},
  eprint = {2610.15950},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.15950},
}
RIS
TY  - UNPB
TI  - Uniform-in-Time and Uniform-in-Temperature Particle Approximations of Coulomb MMD Flows
AU  - Junho Lee
AU  - Promit Ghosal
PY  - 2026
DA  - 2026/10/11
PB  - arXiv
DO  - 10.48550/arxiv.2610.15950
UR  - https://arxiv.org/abs/2610.15950
AB  - We study particle approximations of maximum mean discrepancy (MMD) flows with the Coulomb kernel on the torus 𝕋^d, d≥2, driven by thermal noise of strength τ≥0. For smooth, strictly positive initial and target densities and i.i.d. initial positions, we prove that the expected Wasserstein-1 distance between the empirical measure and the mean-field solution is at most of order N^−1/d for d≥3 and √((log N)/N) for d=2, uniformly in time and in the temperature τ≥0, including the deterministic case τ=0. At each fixed τ>0, we also prove propagation of chaos in total variation for any k-fold marginal distributions of particles, uniformly in time, with a bound of order √(k/τ) times the empirical convergence rate. Our proof uses the modulated free-energy method of Bresch, Jabin, and Wang to reduce the particle convergence estimate to a relaxation estimate that controls the time integral of the mean-field velocity gradient uniformly in τ≥0. Near zero temperature, where the smoothing effect of the noise degenerates, we obtain this estimate by recentering the dynamics at thermal equilibrium and using the damping generated by the Coulomb interaction.
ER  -