pith. sign in

arxiv: 2605.13186 · v4 · pith:YI3YOOBDnew · submitted 2026-05-13 · 🧮 math.AP · math.OC· math.PR

Nesterov acceleration for the Wasserstein minimization of displacement-convex free energies

Pierre Monmarch\'e This is my paper

Pith reviewed 2026-05-22 10:33 UTC · model grok-4.3

classification 🧮 math.AP math.OCmath.PR
keywords Nesterov accelerationWasserstein gradient flowdisplacement convexityunderdamped LangevinVlasov-Fokker-Planck equationPolyak-Lojasiewicz constant
0
0 comments X

The pith

The mean-field underdamped Langevin process achieves Nesterov acceleration for the Wasserstein gradient flow of displacement-convex free energies.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper shows that the mean-field underdamped Langevin process, associated with the nonlinear Vlasov-Fokker-Planck equation, achieves Nesterov acceleration compared to the Wasserstein gradient flow for displacement-convex free energies. The convergence rate is on the order of the square root of the Polyak-Łojasiewicz constant, which is optimal for the gradient flow. This builds upon a recent breakthrough result in the linear case and extends it to the nonlinear setting. A sympathetic reader would care because such accelerations can significantly speed up optimization and sampling procedures in spaces of probability measures.

Core claim

The mean-field underdamped Langevin process associated to the non-linear Vlasov-Fokker-Planck equation achieves a Nesterov acceleration with respect to the Wasserstein gradient flow of a displacement-convex free energy, in the sense that it converges at a rate of order given by the square-root of the Polyak-Łojasiewicz constant of the free energy, which is the optimal convergence rate for the corresponding gradient flow.

What carries the argument

The mean-field underdamped Langevin process and the associated nonlinear Vlasov-Fokker-Planck equation that enable the extension of the diffusive-to-ballistic improvement to the nonlinear case.

If this is right

  • If correct, underdamped dynamics converge faster than overdamped ones in Wasserstein distance for displacement-convex energies.
  • The result applies to a wide class of free energies arising in mean-field limits of interacting particles.
  • It suggests that Nesterov-type accelerations are possible in continuous-time mean-field optimization.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This could lead to new accelerated sampling methods for distributions with displacement-convex potentials.
  • Discrete-time versions of the process might yield practical algorithms with similar rates.
  • The approach may generalize to other metrics or convexity notions in optimal transport.

Load-bearing premise

The recent breakthrough on diffusive-to-ballistic improvement in the linear case extends to the nonlinear displacement-convex setting.

What would settle it

A calculation or simulation for a concrete displacement-convex free energy, such as the entropy plus quadratic potential, where the underdamped process does not exhibit the improved sqrt(PL) convergence rate.

read the original abstract

We show that the mean-field underdamped Langevin process (associated to the non-linear Vlasov-Fokker-Planck equation) achieves a Nesterov acceleration with respect to the Wasserstein gradient flow of a displacement-convex free energy, in the sense that it converges at a rate of order given by the square-root of the Polyak-{\L}ojasiewicz constant of the free energy (which is the optimal convergence rate for the corresponding gradient flow). This result has been made possible by the recent breakthrough [42] by Jianfeng Lu, which establishes such a \emph{diffusive-to-ballistic} improvement in term of entropy in the linear case.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The paper claims that the mean-field underdamped Langevin process associated to the nonlinear Vlasov-Fokker-Planck equation achieves Nesterov acceleration for the Wasserstein gradient flow of displacement-convex free energies, attaining a convergence rate of order sqrt(μ) where μ is the Polyak-Łojasiewicz constant of the free energy. This extends the diffusive-to-ballistic improvement established for the linear case in reference [42].

Significance. If the central extension holds, the result would be significant for connecting hypocoercivity techniques from linear kinetic equations to nonlinear mean-field systems in optimal transport. It offers optimal rates for a broad class of displacement-convex energies relevant to sampling and variational problems, building directly on the recent breakthrough in [42] without introducing new free parameters.

major comments (1)
  1. [§4] §4, proof of Theorem 4.1: The transfer of the sqrt(PL) rate from the linear setting of [42] to the nonlinear Vlasov-Fokker-Planck equation relies on absorbing the interaction term via displacement convexity alone. However, the entropy dissipation or hypocoercivity estimates do not appear to include an explicit uniform bound on the additional commutator or cross terms generated by the nonlinear Fokker-Planck operator when the interaction strength is comparable to the confining potential; this is load-bearing for the claimed rate.
minor comments (2)
  1. The abstract states the result is 'made possible by' [42] but does not outline the key new estimate that controls the nonlinear terms; adding one sentence on the role of displacement convexity in the estimates would improve readability.
  2. [§2] Notation for the mean-field underdamped Langevin process (e.g., the precise form of the interaction force in Eq. (2.3)) should be cross-referenced explicitly when first used in the main theorem statement.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their thorough review and constructive comments. We address the major comment below and will revise the manuscript to make the relevant estimates fully explicit.

read point-by-point responses

  1. Referee: [§4] §4, proof of Theorem 4.1: The transfer of the sqrt(PL) rate from the linear setting of [42] to the nonlinear Vlasov-Fokker-Planck equation relies on absorbing the interaction term via displacement convexity alone. However, the entropy dissipation or hypocoercivity estimates do not appear to include an explicit uniform bound on the additional commutator or cross terms generated by the nonlinear Fokker-Planck operator when the interaction strength is comparable to the confining potential; this is load-bearing for the claimed rate.

    Authors: We thank the referee for this observation. Displacement convexity of the free energy is indeed the key mechanism that absorbs the interaction contribution into the entropy dissipation, allowing the hypocoercivity estimates from the linear case [42] to carry over. Nevertheless, we agree that the commutator and cross terms arising from the nonlinear Fokker-Planck operator should be controlled explicitly and uniformly. In the revised version we will insert a short auxiliary estimate (new Lemma 4.3) that bounds these terms by a multiple of the dissipation functional, with a constant depending only on the displacement-convexity modulus and the PL constant μ; the bound remains valid even when the interaction strength is of the same order as the confining potential. This makes the transfer of the √μ rate fully rigorous without additional smallness assumptions. revision: yes

Circularity Check

0 steps flagged

No circularity; central claim extends independent external result [42]

full rationale

The paper asserts that the mean-field underdamped Langevin dynamics and associated nonlinear Vlasov-Fokker-Planck equation achieve sqrt(PL) convergence for displacement-convex free energies by extending the diffusive-to-ballistic entropy improvement proved in [42] (Jianfeng Lu, different author) from the linear case. The abstract explicitly credits this external breakthrough as what 'has been made possible' the result. No self-definitional steps, fitted inputs renamed as predictions, self-citation load-bearing on the central claim, or ansatz smuggling appear in the provided abstract or derivation outline. The load-bearing step is the transfer of the linear rate via displacement convexity and the mean-field process, which is presented as building on an independent prior theorem rather than reducing to the paper's own inputs or prior self-citations by construction. This qualifies as a normal non-circular extension against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The result depends on displacement convexity of the free energy as a standing assumption and on the validity of the mean-field limit for the underdamped Langevin process; no free parameters or new entities are introduced in the abstract.

axioms (1)
  • domain assumption The free energy is displacement-convex.
    This property is required for the Wasserstein gradient flow to satisfy the Polyak-Łojasiewicz inequality and for the acceleration to apply, as stated in the abstract.

pith-pipeline@v0.9.0 · 5638 in / 1306 out tokens · 52686 ms · 2026-05-22T10:33:15.602421+00:00 · methodology

Review history (4 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

72 extracted references · 72 canonical work pages · 4 internal anchors

  1. [1]

    Vari- ational methods for the kinetic fokker–planck equation.Analysis & PDE, 17(6):1953– 2010, 2024

    Dallas Albritton, Scott Armstrong, Jean-Christophe Mourrat, and Matthew Novack. Vari- ational methods for the kinetic fokker–planck equation.Analysis & PDE, 17(6):1953– 2010, 2024

    work page 1953

  2. [2]

    Shifted composition IV: Underdamped Langevin and numerical discretizations with partial acceleration.arXiv preprint arXiv:2506.23062, 2025

    Jason M Altschuler, Sinho Chewi, and Matthew S Zhang. Shifted composition iv: toward ballistic acceleration for log-concave sampling.arXiv preprint arXiv:2506.23062, 2025

    work page arXiv 2025

  3. [3]

    Springer Science & Business Media, 2005

    Luigi Ambrosio, Nicola Gigli, and Giuseppe Savar´ e.Gradient flows: in metric spaces and in the space of probability measures. Springer Science & Business Media, 2005

    work page 2005

  4. [4]

    Hypocoercivity of piecewise deterministic markov process-monte carlo.The Annals of Applied Probability, 31(5):2478–2517, 2021

    Christophe Andrieu, Alain Durmus, Nikolas N¨ usken, and Julien Roussel. Hypocoercivity of piecewise deterministic markov process-monte carlo.The Annals of Applied Probability, 31(5):2478–2517, 2021. 18

    work page 2021

  5. [5]

    Maximum mean discrep- ancy gradient flow.Advances in neural information processing systems, 32, 2019

    Michael Arbel, Anna Korba, Adil Salim, and Arthur Gretton. Maximum mean discrep- ancy gradient flow.Advances in neural information processing systems, 32, 2019

    work page 2019

  6. [6]

    Springer, Cham, 2014

    Dominique Bakry, Ivan Gentil, and Michel Ledoux.Analysis and geometry of Markov dif- fusion operators, volume 348 ofGrundlehren der Mathematischen Wissenschaften [Fun- damental Principles of Mathematical Sciences]. Springer, Cham, 2014

    work page 2014

  7. [7]

    Gradient flow approach to local mean-field spin systems

    Kaveh Bashiri and Anton Bovier. Gradient flow approach to local mean-field spin systems. Stochastic Processes and their Applications, 130(3):1461–1514, 2020

    work page 2020

  8. [8]

    Bakry–´ emery meet villani.Journal of functional analysis, 273(7):2275– 2291, 2017

    Fabrice Baudoin. Bakry–´ emery meet villani.Journal of functional analysis, 273(7):2275– 2291, 2017

    work page 2017

  9. [9]

    A criterion on the free energy for log-Sobolev inequalities in mean-field particle systems.arXiv e-prints, page arXiv:2503.24372, March 2025

    Roland Bauerschmidt, Thierry Bodineau, and Benoˆ ıt Dagallier. A criterion on the free energy for log-Sobolev inequalities in mean-field particle systems.arXiv e-prints, page arXiv:2503.24372, March 2025

    work page arXiv 2025

  10. [10]

    A piecewise deterministic scaling limit of lifted Metropolis-Hastings in the Curie-Weiss model.Ann

    Joris Bierkens and Gareth Roberts. A piecewise deterministic scaling limit of lifted Metropolis-Hastings in the Curie-Weiss model.Ann. Appl. Probab., 27(2):846–882, 2017

    work page 2017

  11. [11]

    Quantitative stability of constant equilibria in a non-linear alignment model of self-propelled particles

    ´Emeric Bouin and Amic Frouvelle. Quantitative stability of constant equilibria in a non- linear alignment model of self-propelled particles.arXiv preprint arXiv:2604.05927, 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  12. [12]

    Hypocoercivity meets lifts.Kinetic and Related Models, 20(0):34–55, 2026

    Giovanni Brigati, Francis L¨ orler, and Lihan Wang. Hypocoercivity meets lifts.Kinetic and Related Models, 20(0):34–55, 2026

    work page 2026

  13. [13]

    Proximal optimal transport modeling of population dynamics

    Charlotte Bunne, Laetitia Papaxanthos, Andreas Krause, and Marco Cuturi. Proximal optimal transport modeling of population dynamics. InInternational Conference on Artificial Intelligence and Statistics, pages 6511–6528. PMLR, 2022

    work page 2022

  14. [14]

    On explicit l 2-convergence rate estimate for un- derdamped langevin dynamics.Archive for Rational Mechanics and Analysis, 247(5):90, 2023

    Yu Cao, Jianfeng Lu, and Lihan Wang. On explicit l 2-convergence rate estimate for un- derdamped langevin dynamics.Archive for Rational Mechanics and Analysis, 247(5):90, 2023

    work page 2023

  15. [15]

    Springer, 2018

    Ren´ e Carmona, Fran¸ cois Delarue, et al.Probabilistic theory of mean field games with applications I-II, volume 3. Springer, 2018

    work page 2018

  16. [16]

    Convergence to equilibrium in wasser- stein distance for damped euler equations with interaction forces.Communications in Mathematical Physics, 365(1):329–361, 2019

    Jos´ e A Carrillo, Young-Pil Choi, and Oliver Tse. Convergence to equilibrium in wasser- stein distance for damped euler equations with interaction forces.Communications in Mathematical Physics, 365(1):329–361, 2019

    work page 2019

  17. [17]

    Carrillo, Robert J

    Jos´ e A. Carrillo, Robert J. McCann, and C´ edric Villani. Kinetic equilibration rates for granular media and related equations: entropy dissipation and mass transportation estimates.Revista Matem´ atica Iberoamericana, 19(3):971 – 1018, 2003

    work page 2003

  18. [18]

    Entropic multipliers method for langevin diffusion and weighted log sobolev inequalities.Journal of Functional Analysis, 277(11):108288, 2019

    Patrick Cattiaux, Arnaud Guillin, Pierre Monmarch´ e, and Chaoen Zhang. Entropic multipliers method for langevin diffusion and weighted log sobolev inequalities.Journal of Functional Analysis, 277(11):108288, 2019

    work page 2019

  19. [19]

    Uniform-in-time propagation of chaos for kinetic mean field Langevin dynamics.Electronic Journal of Probability, 29(none):1 – 43, 2024

    Fan Chen, Yiqing Lin, Zhenjie Ren, and Songbo Wang. Uniform-in-time propagation of chaos for kinetic mean field Langevin dynamics.Electronic Journal of Probability, 29(none):1 – 43, 2024

    work page 2024

  20. [20]

    Uniform-in-time propagation of chaos for mean field langevin dynamics.arXiv preprint arXiv:2212.03050, 2022

    Fan Chen, Zhenjie Ren, and Songbo Wang. Uniform-in-time propagation of chaos for mean field langevin dynamics.arXiv preprint arXiv:2212.03050, 2022. 19

    work page arXiv 2022

  21. [21]

    Shi Chen, Qin Li, Oliver Tse, and Stephen J. Wright. Accelerating optimization over the space of probability measures.Journal of Machine Learning Research, 26(31):1–40, 2025

    work page 2025

  22. [22]

    Mean-field langevin dynamics : Exponential convergence and annealing

    L´ ena¨ ıc Chizat. Mean-field langevin dynamics : Exponential convergence and annealing. Transactions on Machine Learning Research, 2022

    work page 2022

  23. [23]

    Dalalyan and Lionel Riou-Durand

    Arnak S. Dalalyan and Lionel Riou-Durand. On sampling from a log-concave density using kinetic Langevin diffusions.Bernoulli, 26(3):1956 – 1988, 2020

    work page 1956

  24. [24]

    Mat´ ıas G Delgadino, Rishabh S Gvalani, Grigorios A Pavliotis, and Scott A Smith. Phase transitions, logarithmic sobolev inequalities, and uniform-in-time propagation of chaos for weakly interacting diffusions.Communications in Mathematical Physics, pages 1–49, 2023

    work page 2023

  25. [25]

    Randomized Hamiltonian Monte Carlo as scaling limit of the bouncy particle sampler and dimension-free convergence rates.The Annals of Applied Probability, 31(6):2612 – 2662, 2021

    George Deligiannidis, Daniel Paulin, Alexandre Bouchard-Cˆ ot´ e, and Arnaud Doucet. Randomized Hamiltonian Monte Carlo as scaling limit of the bouncy particle sampler and dimension-free convergence rates.The Annals of Applied Probability, 31(6):2612 – 2662, 2021

    work page 2021

  26. [26]

    Analysis of a nonreversible markov chain sampler.Annals of Applied Probability, pages 726–752, 2000

    Persi Diaconis, Susan Holmes, and Radford M Neal. Analysis of a nonreversible markov chain sampler.Annals of Applied Probability, pages 726–752, 2000

    work page 2000

  27. [27]

    Hypocoercivity for linear kinetic equations conserving mass.Transactions of the American Mathematical Society, 367(6):3807–3828, 2015

    Jean Dolbeault, Cl´ ement Mouhot, and Christian Schmeiser. Hypocoercivity for linear kinetic equations conserving mass.Transactions of the American Mathematical Society, 367(6):3807–3828, 2015

    work page 2015

  28. [28]

    Asymptotic bias of inexact markov chain monte carlo methods in high dimension.The Annals of Applied Probability, 34(4):3435–3468, 2024

    Alain Durmus and Andreas Eberle. Asymptotic bias of inexact markov chain monte carlo methods in high dimension.The Annals of Applied Probability, 34(4):3435–3468, 2024

    work page 2024

  29. [29]

    Non-reversible lifts of reversible diffusion processes and relaxation times

    L¨ orler Francis Eberle, Andreas. Non-reversible lifts of reversible diffusion processes and relaxation times

    work page

  30. [30]

    Continuized accelerations of deterministic and stochastic gradient descents, and of gossip algorithms.Advances in Neural Information Processing Systems, 34:28054–28066, 2021

    Mathieu Even, Rapha¨ el Berthier, Francis Bach, Nicolas Flammarion, Hadrien Hendrikx, Pierre Gaillard, Laurent Massouli´ e, and Adrien Taylor. Continuized accelerations of deterministic and stochastic gradient descents, and of gossip algorithms.Advances in Neural Information Processing Systems, 34:28054–28066, 2021

    work page 2021

  31. [31]

    Sharp hypocoercive convergence estimates for underdamped Langevin dynamics via the modified $L^2$ method

    Zexi Fan, Bowen Li, and Jianfeng Lu. Sharp hypocoercive convergence estimates for underdamped Langevin dynamics via the modifiedL 2 method.arXiv preprint arXiv:2604.10068, 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  32. [32]

    On the rate of convergence in wasserstein distance of the empirical measure.Probability theory and related fields, 162(3-4):707–738, 2015

    Nicolas Fournier and Arnaud Guillin. On the rate of convergence in wasserstein distance of the empirical measure.Probability theory and related fields, 162(3-4):707–738, 2015

    work page 2015

  33. [33]

    Spectral decompositions andL 2-operator norms of toy hypocoercive semi-groups.Kinetic and related models, 6(2):317–372, 2013

    S´ ebastien Gadat and Laurent Miclo. Spectral decompositions andL 2-operator norms of toy hypocoercive semi-groups.Kinetic and related models, 6(2):317–372, 2013

    work page 2013

  34. [34]

    A math- ematical perspective on transformers.Bulletin of the American Mathematical Society, 62(3):427–479, 2025

    Borjan Geshkovski, Cyril Letrouit, Yury Polyanskiy, and Philippe Rigollet. A math- ematical perspective on transformers.Bulletin of the American Mathematical Society, 62(3):427–479, 2025

    work page 2025

  35. [35]

    HMC and Underdamped Langevin United in the Unadjusted Convex Smooth Case.SIAM/ASA Journal on Uncertainty Quantification, 13(1):278–303, 2025

    Nicolas Gouraud, Pierre Le Bris, Adrien Majka, and Pierre Monmarch´ e. HMC and Underdamped Langevin United in the Unadjusted Convex Smooth Case.SIAM/ASA Journal on Uncertainty Quantification, 13(1):278–303, 2025. 20

    work page 2025

  36. [36]

    Uniform Poincar´ e and log- arithmic Sobolev inequalities for mean field particle systems.The Annals of Applied Probability, 32(3):1590 – 1614, 2022

    Arnaud Guillin, Wei Liu, Liming Wu, and Chaoen Zhang. Uniform Poincar´ e and log- arithmic Sobolev inequalities for mean field particle systems.The Annals of Applied Probability, 32(3):1590 – 1614, 2022

    work page 2022

  37. [37]

    Uniform long-time and propagation of chaos estimates for mean field kinetic particles in non-convex landscapes.Journal of Statistical Physics, 185:1–20, 2021

    Arnaud Guillin and Pierre Monmarch´ e. Uniform long-time and propagation of chaos estimates for mean field kinetic particles in non-convex landscapes.Journal of Statistical Physics, 185:1–20, 2021

    work page 2021

  38. [38]

    Short and long time behavior of the fokker–planck equation in a confining potential and applications.Journal of Functional Analysis, 244(1):95–118, 2007

    Fr´ ed´ eric H´ erau. Short and long time behavior of the fokker–planck equation in a confining potential and applications.Journal of Functional Analysis, 244(1):95–118, 2007

    work page 2007

  39. [39]

    Isotropic hypoellipticity and trend to equilibrium for the fokker-planck equation with a high-degree potential.Archive for Rational Mechanics and Analysis, 171(2):151–218, 2004

    Fr´ ed´ eric H´ erau and Francis Nier. Isotropic hypoellipticity and trend to equilibrium for the fokker-planck equation with a high-degree potential.Archive for Rational Mechanics and Analysis, 171(2):151–218, 2004

    work page 2004

  40. [40]

    Mean-field Langevin dy- namics and energy landscape of neural networks.Annales de l’Institut Henri Poincar´ e, Probabilit´ es et Statistiques, 57(4):2043 – 2065, 2021

    Kaitong Hu, Zhenjie Ren, David ˇSiˇ ska, and Lukasz Szpruch. Mean-field Langevin dy- namics and energy landscape of neural networks.Annales de l’Institut Henri Poincar´ e, Probabilit´ es et Statistiques, 57(4):2043 – 2065, 2021

    work page 2043

  41. [41]

    Linear convergence of gradient and proximal-gradient methods under the polyak- lojasiewicz condition

    Hamed Karimi, Julie Nutini, and Mark Schmidt. Linear convergence of gradient and proximal-gradient methods under the polyak- lojasiewicz condition. InJoint European conference on machine learning and knowledge discovery in databases, pages 795–811. Springer, 2016

    work page 2016

  42. [42]

    Variational inference via wasserstein gradient flows.Advances in Neural Information Processing Systems, 35:14434–14447, 2022

    Marc Lambert, Sinho Chewi, Francis Bach, Silv` ere Bonnabel, and Philippe Rigollet. Variational inference via wasserstein gradient flows.Advances in Neural Information Processing Systems, 35:14434–14447, 2022

    work page 2022

  43. [43]

    Convergence rates for an adap- tive biasing potential scheme from a wasserstein optimization perspective.Nonlinearity, 39(4):045016, 2026

    Tony Leli` evre, Xuyang Lin, and Pierre Monmarch´ e. Convergence rates for an adap- tive biasing potential scheme from a wasserstein optimization perspective.Nonlinearity, 39(4):045016, 2026

    work page 2026

  44. [44]

    Partial differential equations and stochastic methods in moleculardynamics.Acta Numerica, 25:681–880, 2016

    Tony Lelievre and Gabriel Stoltz. Partial differential equations and stochastic methods in moleculardynamics.Acta Numerica, 25:681–880, 2016

    work page 2016

  45. [45]

    Ensembles semi-analytiques.IHES notes, page 220, 1965

    Stanislaw Lojasiewicz. Ensembles semi-analytiques.IHES notes, page 220, 1965

    work page 1965

  46. [46]

    A sharp hypocoercive entropy decay estimate for underdamped Langevin dynamics

    Jianfeng Lu. A sharp hypocoercive entropy decay estimate for underdamped Langevin dynamics.arXiv e-prints, page arXiv:2605.01933, May 2026

    work page internal anchor Pith review Pith/arXiv arXiv 2026

  47. [47]

    On explicit l 2-convergence rate estimate for piecewise deterministic markov processes in mcmc algorithms.The Annals of Applied Probability, 32(2):1333–1361, 2022

    Jianfeng Lu and Lihan Wang. On explicit l 2-convergence rate estimate for piecewise deterministic markov processes in mcmc algorithms.The Annals of Applied Probability, 32(2):1333–1361, 2022

    work page 2022

  48. [48]

    A mean field view of the land- scape of two-layer neural networks.Proceedings of the National Academy of Sciences, 115(33):E7665–E7671, 2018

    Song Mei, Andrea Montanari, and Phan-Minh Nguyen. A mean field view of the land- scape of two-layer neural networks.Proceedings of the National Academy of Sciences, 115(33):E7665–E7671, 2018

    work page 2018

  49. [49]

    On the implicit regularization of langevin dynamics with projected noise.arXiv preprint arXiv:2602.12257, 2026

    Govind Menon, Austin J Stromme, and Adrien Vacher. On the implicit regularization of langevin dynamics with projected noise.arXiv preprint arXiv:2602.12257, 2026

    work page arXiv 2026

  50. [50]

    ´Etude spectrale minutieuse de processus moins ind´ ecis que les autres

    Laurent Miclo and Pierre Monmarch´ e. ´Etude spectrale minutieuse de processus moins ind´ ecis que les autres. InS´ eminaire de Probabilit´ es XLV, volume 2078 ofLecture Notes in Math., pages 459–481. Springer, Cham, 2013. 21

    work page 2078

  51. [51]

    Piecewise deterministic simulated annealing.ALEA Lat

    Pierre Monmarch´ e. Piecewise deterministic simulated annealing.ALEA Lat. Am. J. Probab. Math. Stat., 13(1):357–398, 2016

    work page 2016

  52. [52]

    Generalized Γ calculus and application to interacting particles on a graph.Potential Analysis, 50:439–466, 2019

    Pierre Monmarch´ e. Generalized Γ calculus and application to interacting particles on a graph.Potential Analysis, 50:439–466, 2019

    work page 2019

  53. [53]

    An entropic approach for Hamiltonian Monte Carlo: The idealized case.The Annals of Applied Probability, 34(2):2243 – 2293, 2024

    Pierre Monmarch´ e. An entropic approach for Hamiltonian Monte Carlo: The idealized case.The Annals of Applied Probability, 34(2):2243 – 2293, 2024

    work page 2024

  54. [54]

    Free energy Wasserstein gradient flow and their particle counter- parts: toy model, (degenerate) PL inequalities and exit times.arXiv e-prints, page arXiv:2510.16506, October 2025

    Pierre Monmarch´ e. Free energy Wasserstein gradient flow and their particle counter- parts: toy model, (degenerate) PL inequalities and exit times.arXiv e-prints, page arXiv:2510.16506, October 2025

    work page arXiv 2025

  55. [55]

    Uniform log-sobolev inequalities for mean field particles beyond flat- convexity.Stochastic Processes and their Applications, 2025

    Pierre Monmarch´ e. Uniform log-sobolev inequalities for mean field particles beyond flat- convexity.Stochastic Processes and their Applications, 2025

    work page 2025

  56. [56]

    Local convergence rates for wasserstein gradient flows and mckean-vlasov equations with multiple stationary solutions.Probability Theory and Related Fields, pages 1–59, 2025

    Pierre Monmarch´ e and Julien Reygner. Local convergence rates for wasserstein gradient flows and mckean-vlasov equations with multiple stationary solutions.Probability Theory and Related Fields, pages 1–59, 2025

    work page 2025

  57. [57]

    Exact targeting of gibbs distributions using velocity-jump processes.Stochastics and Partial Differential Equa- tions: Analysis and Computations, pages 1–40, 2022

    Pierre Monmarch´ e, Matthias Rousset, and Pierre-Andr´ e Zitt. Exact targeting of gibbs distributions using velocity-jump processes.Stochastics and Partial Differential Equa- tions: Analysis and Computations, pages 1–40, 2022

    work page 2022

  58. [58]

    Long-time behaviour and propagation of chaos for mean field kinetic particles.Stochastic Processes and their Applications, 127(6):1721–1737, 2017

    Pierre Monmarch´ e. Long-time behaviour and propagation of chaos for mean field kinetic particles.Stochastic Processes and their Applications, 127(6):1721–1737, 2017

    work page 2017

  59. [59]

    A method for solving the convex programming problem with convergence rate o (1/k2)

    Yurii Nesterov. A method for solving the convex programming problem with convergence rate o (1/k2). InDokl akad nauk Sssr, volume 269, page 543, 1983

    work page 1983

  60. [60]

    The geometry of dissipative evolution equations: the porous medium equation

    Felix Otto. The geometry of dissipative evolution equations: the porous medium equation. 2001

    work page 2001

  61. [61]

    E. A. J. F. Peters and G. de With. Rejection-free monte carlo sampling for general potentials.Phys. Rev. E 85, 026703, 2012

    work page 2012

  62. [62]

    Entropic approximation of wasserstein gradient flows.SIAM Journal on Imaging Sciences, 8(4):2323–2351, 2015

    Gabriel Peyr´ e. Entropic approximation of wasserstein gradient flows.SIAM Journal on Imaging Sciences, 8(4):2323–2351, 2015

    work page 2015

  63. [63]

    Some methods of speeding up the convergence of iteration methods.Ussr computational mathematics and mathematical physics, 4(5):1–17, 1964

    Boris T Polyak. Some methods of speeding up the convergence of iteration methods.Ussr computational mathematics and mathematical physics, 4(5):1–17, 1964

    work page 1964

  64. [64]

    Etienne Sandier and Sylvia Serfaty. Gamma-convergence of gradient flows with applica- tions to ginzburg-landau.Communications on Pure and Applied Mathematics: A Journal Issued by the Courant Institute of Mathematical Sciences, 57(12):1627–1672, 2004

    work page 2004

  65. [65]

    Understanding the accelera- tion phenomenon via high-resolution differential equations.Mathematical Programming, 195(1):79–148, 2022

    Bin Shi, Simon S Du, Michael I Jordan, and Weijie J Su. Understanding the accelera- tion phenomenon via high-resolution differential equations.Mathematical Programming, 195(1):79–148, 2022

    work page 2022

  66. [66]

    A differential equation for model- ing nesterov’s accelerated gradient method: Theory and insights.Journal of Machine Learning Research, 17(153):1–43, 2016

    Weijie Su, Stephen Boyd, and Emmanuel J Candes. A differential equation for model- ing nesterov’s accelerated gradient method: Theory and insights.Journal of Machine Learning Research, 17(153):1–43, 2016. 22

    work page 2016

  67. [67]

    Quantitative propagation of chaos of mckean-vlasov equations via the master equation

    Alvin Tsz Ho Tse. Quantitative propagation of chaos of mckean-vlasov equations via the master equation. 2019

    work page 2019

  68. [68]

    Hypocoercivity.Mem

    C´ edric Villani. Hypocoercivity.Mem. Amer. Math. Soc., 202(950):iv+141, 2009

    work page 2009

  69. [69]

    Uniform log-Sobolev inequalities for mean field particles with flat-convex energy.arXiv e-prints, page arXiv:2408.03283, August 2024

    Songbo Wang. Uniform log-Sobolev inequalities for mean field particles with flat-convex energy.arXiv e-prints, page arXiv:2408.03283, August 2024

    work page arXiv 2024

  70. [70]

    Large-scale concentration and relaxation for mean-field langevin particle systems.arXiv preprint arXiv:2508.16428, 2025

    Songbo Wang. Large-scale concentration and relaxation for mean-field langevin particle systems.arXiv preprint arXiv:2508.16428, 2025

    work page internal anchor Pith review arXiv 2025

  71. [71]

    Accelerated information gradient flow.Journal of Scientific Computing, 90(1):11, 2022

    Yifei Wang and Wuchen Li. Accelerated information gradient flow.Journal of Scientific Computing, 90(1):11, 2022

    work page 2022

  72. [72]

    A lyapunov analysis of accelerated methods in optimization.Journal of Machine Learning Research, 22(113):1–34, 2021

    Ashia C Wilson, Ben Recht, and Michael I Jordan. A lyapunov analysis of accelerated methods in optimization.Journal of Machine Learning Research, 22(113):1–34, 2021. 23

    work page 2021