pith. sign in

arxiv: 2605.30560 · v1 · pith:YGYIE2W6new · submitted 2026-05-28 · 🧮 math.PR · math.OC

Projected McKean--Vlasov Dynamics for Entropic Weak Optimal Transport

Nathan Sauldubois , Xin Zhang This is my paper

Pith reviewed 2026-06-29 05:19 UTC · model grok-4.3

classification 🧮 math.PR math.OC
keywords weak optimal transportMcKean-Vlasov dynamicsadapted Wasserstein distanceentropic regularizationgradient flowsmartingale optimal transportparticle approximation
0
0 comments X

The pith

A projected McKean-Vlasov dynamics converges in adapted Wasserstein topology to the minimizer of the entropic weak optimal transport problem.

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

The paper derives a coupled McKean-Vlasov SDE by projecting the formal gradient flow of the entropic weak transport cost onto the manifold of couplings with fixed marginals. A key new term averages the conditional weak-transport force at each location to keep the nonlinear dependence on conditional laws intact while enforcing the marginal constraints. Under mild integrability and regularity assumptions the authors prove weak existence and uniqueness in law for the SDE, then show that its law converges in the adapted Wasserstein metric to the unique minimizer of the regularized problem. The construction also supplies a particle approximation that can be simulated on classical and martingale transport examples.

Core claim

From the tangent structure of adapted Wasserstein space and the projection onto couplings with prescribed marginals, a coupled McKean-Vlasov SDE is obtained whose novel projection averages a weak-transport force that already depends on the conditional law of Y given X; under mild assumptions this flow converges in the adapted Wasserstein topology to the unique minimizer of the entropic weak optimal transport problem.

What carries the argument

The projected McKean-Vlasov SDE whose projection term averages the weak-transport force at each Y-location conditional on the law of Y given X, preserving both marginals and the nonlinear weak-transport structure.

If this is right

  • The SDE supplies a dynamical characterization of the entropy-regularized weak transport problem in adapted Wasserstein geometry.
  • Particle approximations of the dynamics yield a numerical method for computing the transport plans.
  • Weak existence and uniqueness in law hold for the projected equation under the stated assumptions.
  • The convergence result applies both to standard optimal transport and to martingale optimal transport examples.

Where Pith is reading between the lines

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

  • The same projection device may extend to other gradient flows whose costs depend nonlinearly on conditional distributions.
  • Adapted Wasserstein dynamics could be used to approximate solutions of related problems such as barycenters subject to martingale constraints.
  • Efficient discretization of the conditional averaging step inside the projection could produce faster algorithms than direct particle simulation.

Load-bearing premise

Mild integrability and regularity assumptions on the cost function and initial data suffice for existence, uniqueness in law, and convergence.

What would settle it

A numerical trajectory of the projected SDE on a low-dimensional Gaussian example whose entropic weak transport minimizer is known in closed form, if it fails to approach that minimizer in adapted Wasserstein distance, would falsify the convergence claim.

Figures

Figures reproduced from arXiv: 2605.30560 by Nathan Sauldubois, Xin Zhang.

Figure 1
Figure 1. Figure 1: Evolution of the empirical coupling for the uniform–Gaussian optimal transport [PITH_FULL_IMAGE:figures/full_fig_p016_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Convergence of the empirical objective value for the uniform–Gaussian optimal [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Evolution of the empirical coupling for the Gaussian–Gaussian martingale opti [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Convergence of the empirical objective value for the Gaussian–Gaussian mar [PITH_FULL_IMAGE:figures/full_fig_p017_4.png] view at source ↗
read the original abstract

Unlike classical optimal transport, weak transport costs depend nonlinearly on the conditional law of couplings. This feature is essential in problems involving barycenter, conditional moments, and martingale-type constraints. Meanwhile, such conditional dependence makes ordinary Wasserstein geometry insufficient and calls instead for an adapted Wasserstein viewpoint. In this paper, we investigate the entropy-regularized weak optimal transport via gradient flows in adapted Wasserstein space. We derive, from the formal tangent structure of adapted Wasserstein space and the projection onto the set of couplings with prescribed marginals, a coupled McKean--Vlasov SDE. A novel and subtle term is a projection that, at each $Y$-location, averages a weak-transport force that already depends on the conditional law of $Y$ given $X$, thereby preserving marginals while retaining the nonlinear weak-transport structure. Under mild integrability and regularity assumptions, we prove weak existence and uniqueness in law for this projected McKean--Vlasov equation. We then prove that the flow converges, in the adapted Wasserstein topology, to the unique minimizer of the entropic weak optimal transport problem. We also describe a particle approximation and illustrate the dynamics on optimal transport and martingale optimal transport examples.

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

0 major / 3 minor

Summary. The manuscript derives a projected McKean-Vlasov SDE from the formal tangent structure of adapted Wasserstein space and a projection onto couplings with prescribed marginals. Under mild integrability and regularity assumptions it establishes weak existence and uniqueness in law for the SDE, then proves that the resulting flow converges in the adapted Wasserstein topology to the unique minimizer of the entropic weak optimal transport problem. A particle approximation is described and the dynamics are illustrated on classical and martingale optimal transport examples.

Significance. If the well-posedness and convergence results hold, the work supplies a dynamical characterization of entropic weak OT via adapted Wasserstein gradient flows. This is potentially significant for problems with nonlinear conditional dependence, such as barycenters and martingale constraints, where standard Wasserstein geometry is insufficient. The explicit construction of the projection term that averages the conditional weak-transport force while preserving marginals is a technical contribution.

minor comments (3)
  1. [Abstract] The abstract states that the projection 'averages a weak-transport force that already depends on the conditional law of Y given X'; an explicit formula for this averaging operator (perhaps in the form of an integral against the conditional measure) would improve readability before the full derivation appears.
  2. Notation for the adapted Wasserstein distance and the tangent space is used from the outset; a short preliminary subsection recalling the relevant definitions and the precise form of the projection would aid readers unfamiliar with the adapted setting.
  3. The particle approximation is described but not claimed to be rigorously proved; if the authors intend this only as a numerical illustration, a brief remark clarifying its status relative to the main convergence theorem would prevent misinterpretation.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive summary, recognition of the technical contribution, and recommendation of minor revision. No major comments were listed in the report.

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained from tangent structure and standard well-posedness

full rationale

The paper derives the projected McKean-Vlasov SDE formally from the tangent structure of adapted Wasserstein space plus a marginal-preserving projection step. It then states and proves weak existence/uniqueness in law under explicit mild integrability/regularity assumptions, followed by a separate convergence argument to the entropic weak OT minimizer in adapted Wasserstein topology. No quoted step reduces the target minimizer or the flow equation to a fitted parameter or self-citation by construction; the projection is introduced to enforce the marginal constraint while retaining the nonlinear conditional structure. The particle approximation is described but not asserted as a rigorous theorem. This matches the default case of a self-contained derivation with no load-bearing circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claims rest on standard background results in stochastic analysis and optimal transport geometry; no free parameters or invented entities are introduced in the abstract.

axioms (2)
  • domain assumption Existence of a tangent structure on adapted Wasserstein space allowing formal gradient flow derivation
    Invoked to obtain the McKean-Vlasov SDE from the entropy-regularized functional.
  • domain assumption Mild integrability and regularity conditions suffice for weak existence and uniqueness of the projected SDE
    Stated explicitly as the hypothesis under which the existence/uniqueness and convergence theorems hold.

pith-pipeline@v0.9.1-grok · 5749 in / 1409 out tokens · 21687 ms · 2026-06-29T05:19:42.118565+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

55 extracted references · 11 canonical work pages · 2 internal anchors

  1. [1]

    Acciaio, J

    B. Acciaio, J. Backhoff-Veraguas, and A. Zalashko , Causal optimal transport and its links to enlargement of filtrations and continuous-time stochastic optimization , Stochastic Processes and their Applications, 130 (2020), pp. 2918--2953

    2020

  2. [2]

    Acciaio, D

    B. Acciaio, D. Kr s ek, G. Pammer, and M. Rodrigues , Absolutely continuous curves of stochastic processes , arXiv preprint arXiv:2506.13634, (2025)

    work page arXiv 2025

  3. [3]

    Ambrosio, N

    L. Ambrosio, N. Gigli, and G. Savar \'e , Gradient Flows: In Metric Spaces and in the Space of Probability Measures , Birkh \"a user, Basel, 2 ed., 2008

    2008

  4. [4]

    Backhoff, D

    J. Backhoff, D. Bartl, M. Beiglb \"o ck, and J. Wiesel , Estimating processes in adapted Wasserstein distance , The Annals of Applied Probability, 32 (2022), pp. 529--550

    2022

  5. [5]

    Backhoff-Veraguas, D

    J. Backhoff-Veraguas, D. Bartl, M. Beiglb \"o ck, and M. Eder , Adapted Wasserstein distances and stability in mathematical finance , Finance and Stochastics, 24 (2020), pp. 601--632

    2020

  6. [6]

    o ck, M. Huesmann, and S. K \

    J. Backhoff-Veraguas, M. Beiglb \"o ck, M. Huesmann, and S. K \"a llblad , Martingale Benamou--Brenier : A probabilistic perspective , The Annals of Probability, 48 (2020), pp. 2258--2289

    2020

  7. [7]

    Backhoff-Veraguas, M

    J. Backhoff-Veraguas, M. Beiglb \"o ck, Y. Lin, and A. Zalashko , Causal transport in discrete time and applications , SIAM Journal on Optimization, 27 (2017), pp. 2528--2562

    2017

  8. [8]

    Backhoff-Veraguas, M

    J. Backhoff-Veraguas, M. Beiglb \"o ck, and G. Pammer , Existence, duality, and cyclical monotonicity for weak transport costs , Calculus of Variations and Partial Differential Equations, 58 (2019), p. 203

    2019

  9. [9]

    J. D. Backhoff-Veraguas and G. Pammer , Applications of weak transport theory , Bernoulli, 28 (2022), pp. 370--394

    2022

  10. [10]

    721--752

    height 2pt depth -1.6pt width 23pt, Stability of martingale optimal transport and weak optimal transport , The Annals of Applied Probability, 32 (2022), pp. 721--752

    2022

  11. [11]

    Bartl, M

    D. Bartl, M. Beiglb \"o ck, and G. Pammer , The Wasserstein space of stochastic processes , Journal of the European Mathematical Society, 28 (2026), pp. 393--454

    2026

  12. [12]

    Bartl, M

    D. Bartl, M. Beiglb \"o ck, G. Pammer, S. Schrott, and X. Zhang , The Wasserstein space of stochastic processes in continuous time , arXiv preprint arXiv:2501.14135, (2025)

    work page arXiv 2025

  13. [13]

    Bartl, S

    D. Bartl, S. Drapeau, J. Ob \'o j, and J. Wiesel , Sensitivity analysis of Wasserstein distributionally robust optimization problems , Proceedings of the Royal Society A, 477 (2021), p. 20210176

    2021

  14. [14]

    Bartl and J

    D. Bartl and J. Wiesel , Sensitivity of multiperiod optimization problems with respect to the adapted wasserstein distance , SIAM Journal on Financial Mathematics, 14 (2023), pp. 704--720

    2023

  15. [15]

    Beiglb \"o ck, P

    M. Beiglb \"o ck, P. Henry-Labord \`e re, and F. Penkner , Model-independent bounds for option prices: A mass transport approach , Finance and Stochastics, 17 (2013), pp. 477--501

    2013

  16. [16]

    Beiglb \"o ck, B

    M. Beiglb \"o ck, B. Jourdain, W. Margheriti, and G. Pammer , Stability of the weak martingale optimal transport problem , The Annals of Applied Probability, 33 (2023), pp. 5382--5412

    2023

  17. [17]

    Beiglb \"o ck, G

    M. Beiglb \"o ck, G. Pammer, L. Riess, and S. Schrott , The fundamental theorem of weak optimal transport , arXiv preprint arXiv:2501.16316, (2025)

    work page arXiv 2025

  18. [18]

    Beiglb \"o ck, G

    M. Beiglb \"o ck, G. Pammer, S. Schrott, and X. Zhang , Representing general stochastic processes as martingale laws , arXiv preprint arXiv:2312.16725, (2023)

    work page arXiv 2023

  19. [19]

    V. I. Bogachev, N. V. Krylov, M. R \"o ckner, and S. V. Shaposhnikov , Fokker--Planck--Kolmogorov Equations , vol. 207 of Mathematical Surveys and Monographs, American Mathematical Society, Providence, RI, 2022

    2022

  20. [20]

    Bolley and C

    F. Bolley and C. Villani , Weighted csisz \'a r--kullback--pinsker inequalities and applications to transportation inequalities , in Annales de la Facult \'e des Sciences de Toulouse: Math \'e matiques, vol. 14, 2005, pp. 331--352

    2005

  21. [21]

    Br \'e zis , Functional Analysis, Sobolev Spaces and Partial Differential Equations , Universitext, Springer, New York, 2011

    H. Br \'e zis , Functional Analysis, Sobolev Spaces and Partial Differential Equations , Universitext, Springer, New York, 2011

    2011

  22. [22]

    Brunick and S

    G. Brunick and S. Shreve , Mimicking an it \^o process by a solution of a stochastic differential equation , The Annals of Applied Probability, 23 (2013), pp. 1584--1628

    2013

  23. [23]

    Carlier, H

    G. Carlier, H. Malamut, and M. Sylvestre , Weak optimal transport with moment constraints: Constraint qualification, dual attainment and entropic regularization , arXiv preprint arXiv:2511.16211, (2025)

    work page arXiv 2025

  24. [24]

    F. Chen, G. Conforti, Z. Ren, and X. Wang , Convergence of sinkhorn's algorithm for entropic martingale optimal transport problem , Mathematics of Operations Research, (2026). Articles in Advance

    2026

  25. [25]

    Y. Chen, T. T. Georgiou, and M. Pavon , On the relation between optimal transport and schr \"o dinger bridges: A stochastic control viewpoint , Journal of Optimization Theory and Applications, 169 (2016), pp. 671--691

    2016

  26. [26]

    Conforti, D

    G. Conforti, D. Lacker, and S. Pal , Projected langevin dynamics and a gradient flow for entropic optimal transport , Journal of the European Mathematical Society, (2025). Published online first

    2025

  27. [27]

    Cuturi , Sinkhorn distances: Lightspeed computation of optimal transport , in Advances in Neural Information Processing Systems, vol

    M. Cuturi , Sinkhorn distances: Lightspeed computation of optimal transport , in Advances in Neural Information Processing Systems, vol. 26, 2013, pp. 2292--2300

    2013

  28. [28]

    Eckstein and G

    S. Eckstein and G. Pammer , Computational methods for adapted optimal transport , The Annals of Applied Probability, 34 (2024), pp. 675--713

    2024

  29. [29]

    Compactness in Adapted Weak Topologies

    M. Eder , Compactness in adapted weak topologies , arXiv preprint arXiv:1905.00856, (2019)

    work page internal anchor Pith review Pith/arXiv arXiv 1905

  30. [30]

    Gozlan and N

    N. Gozlan and N. Juillet , On a mixture of brenier and strassen theorems , Proceedings of the London Mathematical Society, 120 (2020), pp. 434--463

    2020

  31. [31]

    Gozlan, C

    N. Gozlan, C. Roberto, P.-M. Samson, and P. Tetali , Kantorovich duality for general transport costs and applications , Journal of Functional Analysis, 273 (2017), pp. 3327--3405

    2017

  32. [32]

    Henry-Labord \`e re , Model-Free Hedging: A Martingale Optimal Transport Viewpoint , Chapman and Hall/CRC Financial Mathematics Series, CRC Press, Boca Raton, FL, 2017

    P. Henry-Labord \`e re , Model-Free Hedging: A Martingale Optimal Transport Viewpoint , Chapman and Hall/CRC Financial Mathematics Series, CRC Press, Boca Raton, FL, 2017

    2017

  33. [33]

    Hern \'a ndez and L

    C. Hern \'a ndez and L. Tangpi , Marginal flows of non-entropic weak schr " odinger bridges , arXiv preprint arXiv:2512.21261, (2025)

    work page arXiv 2025

  34. [34]

    Huesmann and D

    M. Huesmann and D. Trevisan , A Benamou--Brenier formulation of martingale optimal transport , Bernoulli, 25 (2019), pp. 2729--2757

    2019

  35. [35]

    Jiang and J

    Y. Jiang and J. Ob \'o j , Sensitivity of causal distributionally robust optimization , arXiv preprint arXiv:2408.17109, (2024)

    work page arXiv 2024

  36. [36]

    Jordan, D

    R. Jordan, D. Kinderlehrer, and F. Otto , The variational formulation of the fokker--planck equation , SIAM Journal on Mathematical Analysis, 29 (1998), pp. 1--17

    1998

  37. [37]

    Kr s ek and G

    D. Kr s ek and G. Pammer , General duality and dual attainment for adapted transport , Applied Mathematics & Optimization, 91 (2025), p. 52

    2025

  38. [38]

    Lacker, M

    D. Lacker, M. Shkolnikov, and J. Zhang , Superposition and mimicking theorems for conditional mckean--vlasov equations , Journal of the European Mathematical Society, 25 (2023), pp. 3229--3288

    2023

  39. [39]

    Causal transference plans and their Monge-Kantorovich problems

    R. Lassalle , Causal transference plans and their monge--kantorovich problems , arXiv preprint arXiv:1303.6925, (2013)

    work page internal anchor Pith review Pith/arXiv arXiv 2013

  40. [40]

    L \'e onard , A survey of the schr \"o dinger problem and some of its connections with optimal transport , Discrete and Continuous Dynamical Systems -- Series A, 34 (2014), pp

    C. L \'e onard , A survey of the schr \"o dinger problem and some of its connections with optimal transport , Discrete and Continuous Dynamical Systems -- Series A, 34 (2014), pp. 1533--1574

    2014

  41. [41]

    Mikami and M

    T. Mikami and M. Thieullen , Optimal transportation problem by stochastic optimal control , SIAM Journal on Control and Optimization, 47 (2008), pp. 1127--1139

    2008

  42. [42]

    Nutz and J

    M. Nutz and J. Wiesel , On the martingale schr \"o dinger bridge between two distributions , arXiv preprint arXiv:2401.05209, (2024)

    work page arXiv 2024

  43. [43]

    Otto , The geometry of dissipative evolution equations: The porous medium equation , Communications in Partial Differential Equations, 26 (2001), pp

    F. Otto , The geometry of dissipative evolution equations: The porous medium equation , Communications in Partial Differential Equations, 26 (2001), pp. 101--174

    2001

  44. [44]

    Pammer , A note on the adapted weak topology in discrete time , Electronic Communications in Probability, 29 (2024), pp

    G. Pammer , A note on the adapted weak topology in discrete time , Electronic Communications in Probability, 29 (2024), pp. 1--13

    2024

  45. [45]

    Peyr \'e and M

    G. Peyr \'e and M. Cuturi , Computational optimal transport with applications to data science , Foundations and Trends in Machine Learning, 11 (2019), pp. 355--607

    2019

  46. [46]

    G. C. Pflug and A. Pichler , A distance for multistage stochastic optimization models , SIAM Journal on Optimization, 22 (2012), pp. 1--23

    2012

  47. [47]

    height 2pt depth -1.6pt width 23pt, Multistage Stochastic Optimization , Springer Series in Operations Research and Financial Engineering, Springer, Cham, 2014

    2014

  48. [48]

    Santambrogio , Optimal Transport for Applied Mathematicians: Calculus of Variations, PDEs, and Modeling , vol

    F. Santambrogio , Optimal Transport for Applied Mathematicians: Calculus of Variations, PDEs, and Modeling , vol. 87 of Progress in Nonlinear Differential Equations and Their Applications, Birkh \"a user, Cham, 2015

    2015

  49. [49]

    height 2pt depth -1.6pt width 23pt, Euclidean, metric, and wasserstein gradient flows: An overview , Bulletin of Mathematical Sciences, 7 (2017), pp. 87--154

    2017

  50. [50]

    Sauldubois , Model risk static-hedging: A constrained distributionally robust optimization approach

    N. Sauldubois , Model risk static-hedging: A constrained distributionally robust optimization approach . Working paper, 2026

    2026

  51. [51]

    Sauldubois and N

    N. Sauldubois and N. Touzi , First order martingale model risk and semi-static hedging , arXiv preprint arXiv:2410.06906, (2024)

    work page arXiv 2024

  52. [52]

    o dinger , \

    E. Schr \"o dinger , \"U ber die umkehrung der naturgesetze , Sitzungsberichte der Preussischen Akademie der Wissenschaften, Physikalisch-Mathematische Klasse, (1931), pp. 144--153

    1931

  53. [53]

    Shu , From hopf--lax formula to optimal weak transfer plan , SIAM Journal on Mathematical Analysis, 52 (2020), pp

    Y. Shu , From hopf--lax formula to optimal weak transfer plan , SIAM Journal on Mathematical Analysis, 52 (2020), pp. 3052--3072

    2020

  54. [54]

    Villani , Topics in Optimal Transportation , vol

    C. Villani , Topics in Optimal Transportation , vol. 58 of Graduate Studies in Mathematics, American Mathematical Society, Providence, RI, 2003

    2003

  55. [55]

    338 of Grundlehren der mathematischen Wissenschaften, Springer, Berlin, 2009

    height 2pt depth -1.6pt width 23pt, Optimal Transport : Old and New , vol. 338 of Grundlehren der mathematischen Wissenschaften, Springer, Berlin, 2009

    2009