pith. sign in

arxiv: 2606.04377 · v1 · pith:NWTECYDWnew · submitted 2026-06-03 · 🧮 math.AP · math.OC· math.PR

A comparison principle for Wasserstein PDEs with state- and law-dependent common noise

Erhan Bayraktar , Ibrahim Ekren , Xihao He , Xin Zhang This is my paper

Pith reviewed 2026-06-28 05:53 UTC · model grok-4.3

classification 🧮 math.AP math.OCmath.PR
keywords Wasserstein spacecomparison principleviscosity solutionsHamilton-Jacobi-Bellman equationscommon noisestochastic filteringLamperti transformmean-field control
0
0 comments X

The pith

A nonlinear flow of measures transforms Wasserstein-space HJB equations with state- and law-dependent common noise into an augmented equation on which a Crandall-Ishii comparison holds.

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

The paper proves a comparison principle for second-order Hamilton-Jacobi-Bellman equations on the Wasserstein space in which the second-order term arises from a general common-noise Hessian that depends on both the state and the current measure. The key step is the construction of a nonlinear flow of measures that augments the state space with an auxiliary variable, converting the common-noise direction into an ordinary second derivative. Under estimates on this flow and structural conditions on the resulting Hamiltonian, the Crandall-Ishii lemma yields uniqueness for semicontinuous viscosity sub- and supersolutions. This framework is applied to identify the value function of a controlled stochastic filtering problem as the unique viscosity solution of its dynamic programming equation.

Core claim

Under the spatial, measure-derivative and negative-Sobolev estimates satisfied by the nonlinear flow of measures, together with structural assumptions on the transformed Hamiltonian, any semicontinuous viscosity subsolution is bounded above by any semicontinuous viscosity supersolution; the same change-of-variable construction also applies to Zakai-type Kolmogorov equations on spaces of finite positive measures.

What carries the argument

The nonlinear flow of measures, which serves as a measure-dependent Lamperti transform that augments the domain to [0,T]×P₂(ℝ)×ℝ and converts the general common-noise Hessian into an ordinary second derivative in the auxiliary variable.

If this is right

  • The value function of any controlled stochastic filtering problem with state- and law-dependent common noise is the unique viscosity solution of its dynamic programming equation.
  • Uniqueness holds for a new class of second-order PDEs on spaces of measures whose common-noise directions depend on both state and conditional law.
  • The same transformation yields comparison principles for Zakai-type Kolmogorov equations on spaces of finite positive measures.
  • Viscosity solutions can be compared directly without requiring the common-noise coefficient to be nondegenerate or independent of the measure.

Where Pith is reading between the lines

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

  • The transformed equation on the augmented space may admit standard numerical schemes that can be pulled back to approximate the original Wasserstein PDE.
  • The approach could be tested on mean-field games whose common noise is generated by a finite-dimensional observation process.
  • If the estimates on the flow can be verified for a broader family of coefficients, the comparison principle would extend to controlled McKean-Vlasov equations with law-dependent volatility.

Load-bearing premise

The nonlinear flow of measures must admit the spatial, measure-derivative and negative-Sobolev estimates needed for the viscosity argument to close.

What would settle it

An explicit pair of distinct semicontinuous viscosity sub- and supersolutions for a concrete filtering equation whose common-noise coefficient satisfies the paper's structural conditions but for which the required estimates on the flow fail.

read the original abstract

We prove a comparison principle for a class of second-order Hamilton--Jacobi--Bellman equations on the Wasserstein space whose second-order term is generated by a general common-noise Hessian. The main difficulty is that the relevant second-order direction is induced by a state- and measure-dependent coefficient, so the associated perturbation of the measure is no longer a translation or a fixed state-dependent transformation. We introduce a nonlinear flow of measures and use it to transform the Wasserstein-space equation into an augmented equation on $[0,T]\times \mathcal P_2(\mathbb R)\times\mathbb R$, where the general Hessian becomes an ordinary second derivative in the auxiliary variable. The construction may be viewed as a measure-dependent Lamperti transform: it removes the common-noise direction at the level of the equation, but unlike the classical one-dimensional Lamperti transform it permits degeneracy of the coefficient and dependence on the conditional law. We establish the spatial, measure-derivative, and negative-Sobolev estimates for this flow that are needed in the viscosity argument. Under structural assumptions on the transformed Hamiltonian, these estimates yield a Crandall--Ishii type comparison theorem for semicontinuous viscosity sub- and supersolutions. This gives, to the best of our knowledge, the first viscosity comparison framework of this kind for the filtering-driven equations considered here, and opens a new class of second-order PDEs on spaces of measures with state- and law-dependent common-noise directions. As an application, we identify the value function of a controlled stochastic filtering problem with state- and law-dependent common noise as the unique viscosity solution of its dynamic programming equation. We also explain how the same change-of-variable viewpoint applies to Zakai-type Kolmogorov equations on spaces of finite positive measures.

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 manuscript proves a comparison principle for second-order Hamilton–Jacobi–Bellman equations on the Wasserstein space whose second-order term arises from a state- and law-dependent common-noise Hessian. The strategy introduces a nonlinear flow of measures that serves as a measure-dependent Lamperti transform, converting the original equation into an augmented PDE on [0,T]×P₂(ℝ)×ℝ in which the common-noise direction becomes an ordinary second derivative. Spatial, measure-derivative, and negative-Sobolev estimates are established for this flow; under structural assumptions on the transformed Hamiltonian these estimates permit a Crandall–Ishii doubling-variables argument that yields the comparison for semicontinuous viscosity sub- and supersolutions. The result is applied to identify the value function of a controlled stochastic filtering problem as the unique viscosity solution of its dynamic-programming equation, and the same viewpoint is indicated for Zakai-type Kolmogorov equations on finite positive measures.

Significance. If the estimates close, the work supplies the first viscosity comparison framework for filtering-driven equations with genuinely law-dependent common noise on Wasserstein space. It thereby opens a new class of second-order PDEs on spaces of measures whose coefficients may degenerate and depend on the conditional law, while also furnishing an explicit identification of the value function for the associated stochastic control problem.

major comments (1)
  1. [paragraph on construction and estimates (abstract); the section deriving the negative-Sobolev bound for the nonlinear fl] The negative-Sobolev estimates for the nonlinear flow (the step that removes the common-noise Hessian) are load-bearing for the entire comparison. Because the flow coefficient is both state- and law-dependent and may degenerate, the usual translation or fixed-map arguments no longer apply; the manuscript must therefore derive a new a-priori bound that remains uniform under law dependence. If this bound fails to close, the transformed equation cannot be fed into the Crandall–Ishii lemma after the change of variables.
minor comments (2)
  1. State the precise structural assumptions imposed on the transformed Hamiltonian (growth, monotonicity, continuity in the measure variable) that are needed for the Crandall–Ishii argument to close; these assumptions should be listed explicitly rather than left implicit.
  2. Clarify whether the spatial and measure-derivative estimates for the flow are obtained under the same degeneracy conditions that are allowed for the negative-Sobolev estimate, or whether additional non-degeneracy is tacitly used.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the positive assessment and for highlighting the central role of the negative-Sobolev estimates. We address the major comment below.

read point-by-point responses

  1. Referee: The negative-Sobolev estimates for the nonlinear flow (the step that removes the common-noise Hessian) are load-bearing for the entire comparison. Because the flow coefficient is both state- and law-dependent and may degenerate, the usual translation or fixed-map arguments no longer apply; the manuscript must therefore derive a new a-priori bound that remains uniform under law dependence. If this bound fails to close, the transformed equation cannot be fed into the Crandall–Ishii lemma after the change of variables.

    Authors: We agree that these estimates are load-bearing. The manuscript constructs the nonlinear flow of measures as a measure-dependent Lamperti transform and derives the required spatial, measure-derivative, and negative-Sobolev bounds in Sections 3–4. The a-priori bound is obtained via energy methods adapted to the Wasserstein metric that remain uniform under both state and law dependence of the coefficient, without invoking translations or fixed maps; the estimates explicitly accommodate degeneracy. These bounds ensure the transformed PDE on [0,T]×P₂(ℝ)×ℝ satisfies the structural hypotheses needed for the Crandall–Ishii argument, closing the comparison. revision: no

Circularity Check

0 steps flagged

Direct proof construction with no circular reductions

full rationale

The paper constructs a nonlinear flow of measures to transform the Wasserstein-space HJB equation into an augmented PDE on [0,T]×P₂(ℝ)×ℝ, derives the necessary spatial/measure-derivative/negative-Sobolev estimates for this flow, and invokes the Crandall-Ishii lemma on the transformed Hamiltonian. No step reduces by construction to a fitted parameter, self-referential definition, or load-bearing self-citation; the estimates are proved directly for the state- and law-dependent coefficient, and the comparison follows from the structural assumptions without circular closure. The argument is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on standard background results in viscosity solution theory for Wasserstein-space PDEs and on the existence of a sufficiently regular nonlinear flow; no free parameters are introduced and the only invented object is the flow itself, which has no independent evidence outside the paper.

axioms (1)
  • standard math Standard properties of the Wasserstein space P_2(R) and the theory of viscosity solutions for second-order PDEs on infinite-dimensional spaces hold.
    Invoked throughout the comparison argument and the application of the Crandall-Ishii lemma.
invented entities (1)
  • nonlinear flow of measures no independent evidence
    purpose: Transforms the Wasserstein-space HJB equation so that the general common-noise Hessian becomes an ordinary second derivative in an auxiliary variable.
    Newly constructed object whose estimates are central to the proof; no external verification supplied.

pith-pipeline@v0.9.1-grok · 5863 in / 1434 out tokens · 38083 ms · 2026-06-28T05:53:02.284532+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

44 extracted references · 4 canonical work pages

  1. [1]

    Adams and Lars Inge Hedberg.Function Spaces and Potential Theory, volume 314 of Grundlehren der mathematischen Wissenschaften

    David R. Adams and Lars Inge Hedberg.Function Spaces and Potential Theory, volume 314 of Grundlehren der mathematischen Wissenschaften. Springer, Berlin, 1996

    1996

  2. [2]

    Springer, New York, 2009

    Alan Bain and Dan Crisan.Fundamentals of Stochastic Filtering, volume 60 ofStochastic Modelling and Applied Probability. Springer, New York, 2009

    2009

  3. [3]

    Randomized filtering and Bellman equation in Wasserstein space for partial observation control problem.Stochastic Process

    Elena Bandini, Andrea Cosso, Marco Fuhrman, and Huyên Pham. Randomized filtering and Bellman equation in Wasserstein space for partial observation control problem.Stochastic Process. Appl., 129(2):674–711, 2019

    2019

  4. [4]

    Viscosity Solutions of Fully Second-Order HJB Equations in the Wasserstein Space.SIAM J

    Erhan Bayraktar, Hang Cheung, Ibrahim Ekren, Jinniao Qiu, Ho Man Tai, and Xin Zhang. Viscosity Solutions of Fully Second-Order HJB Equations in the Wasserstein Space.SIAM J. Control Optim.,

  5. [5]

    Randomized dynamic programming principle and Feynman–Kac representation for optimal control of McKean–Vlasov dynamics.Trans

    Erhan Bayraktar, Andrea Cosso, and Huyên Pham. Randomized dynamic programming principle and Feynman–Kac representation for optimal control of McKean–Vlasov dynamics.Trans. Amer. Math. Soc., 370(3):2115–2160, 2018

    2018

  6. [6]

    Comparison for semi-continuous viscos- ity solutions for second-order PDEs on the Wasserstein space.J

    Erhan Bayraktar, Ibrahim Ekren, Xihao He, and Xin Zhang. Comparison for semi-continuous viscos- ity solutions for second-order PDEs on the Wasserstein space.J. Differential Equations, 455:113963, 2026

    2026

  7. [7]

    A PDE approach for regret bounds under partial monitoring.J

    Erhan Bayraktar, Ibrahim Ekren, and Xin Zhang. A PDE approach for regret bounds under partial monitoring.J. Mach. Learn. Res., 24(299):1–24, 2023

    2023

  8. [8]

    A Smooth Variational Principle on Wasserstein Space.Proc

    Erhan Bayraktar, Ibrahim Ekren, and Xin Zhang. A Smooth Variational Principle on Wasserstein Space.Proc. Amer. Math. Soc., 151(9):4089–4098, 2023

    2023

  9. [9]

    Comparison of viscosity solutions for a class of second-order PDEs on the Wasserstein space.Communications in Partial Differential Equations, 50(4):570–613, 2025

    Erhan Bayraktar, Ibrahim Ekren, and Xin Zhang. Comparison of viscosity solutions for a class of second-order PDEs on the Wasserstein space.Communications in Partial Differential Equations, 50(4):570–613, 2025

    2025

  10. [10]

    Convergence Rate of Particle System for Second- Order PDEs on Wasserstein Space.SIAM J

    Erhan Bayraktar, Ibrahim Ekren, and Xin Zhang. Convergence Rate of Particle System for Second- Order PDEs on Wasserstein Space.SIAM J. Control Optim., 63(3):1768–1782, 2025. 26

    2025

  11. [11]

    Uniform-in-Time Weak Propagation of Chaos for Consensus-Based Optimization.Ann

    Erhan Bayraktar, Ibrahim Ekren, and Hongyi Zhou. Uniform-in-Time Weak Propagation of Chaos for Consensus-Based Optimization.Ann. Appl. Probab., 2025. to appear, arXiv:2502.00582

    work page arXiv 2025

  12. [12]

    Multiplication in Sobolev spaces, revisited.Ark

    Ali Behzadan and Michael Holst. Multiplication in Sobolev spaces, revisited.Ark. Mat., 59(2):275– 306, 2021

    2021

  13. [13]

    Springer, 2013

    Alain Bensoussan, Jens Frehse, and Phillip Yam.Mean Field Games and Mean Field Type Control Theory, volume 101. Springer, 2013

    2013

  14. [14]

    Stochastic Optimal Transport and Hamilton–Jacobi–Bellman Equations on the Set of Probability Measures.Ann

    Charles Bertucci. Stochastic Optimal Transport and Hamilton–Jacobi–Bellman Equations on the Set of Probability Measures.Ann. Inst. H. Poincaré C Anal. Non Linéaire, 42(6):1543–1600, 2025

    2025

  15. [15]

    An Approximation of the Squared Wasserstein Distance and an Application to Hamilton–Jacobi Equations.arXiv preprint arXiv:2409.11793, 2024

    Charles Bertucci and Pierre-Louis Lions. An Approximation of the Squared Wasserstein Distance and an Application to Hamilton–Jacobi Equations.arXiv preprint arXiv:2409.11793, 2024

    work page arXiv 2024

  16. [16]

    Mean-field stochastic differential equations and associated PDEs.Ann

    Rainer Buckdahn, Juan Li, Shige Peng, and Catherine Rainer. Mean-field stochastic differential equations and associated PDEs.Ann. Probab., 45(2):824–878, 2017

    2017

  17. [17]

    Max Reppen, and H

    Matteo Burzoni, Vincenzo Ignazio, A. Max Reppen, and H. Mete Soner. Viscosity Solutions for Controlled McKean–Vlasov Jump-Diffusions.SIAM J. Control Optim., 58(3):1676–1699, 2020

    2020

  18. [18]

    Princeton University Press, 2019

    Pierre Cardaliaguet, François Delarue, Jean-Michel Lasry, and Pierre-Louis Lions.The Master Equation and the Convergence Problem in Mean Field Games, volume 201 ofAnnals of Mathematics Studies. Princeton University Press, 2019

    2019

  19. [19]

    Probability Theory and Stochastic Modelling

    René Carmona and François Delarue.Probabilistic Theory of Mean Field Games with Applications I: Mean Field FBSDEs, Control, and Games. Probability Theory and Stochastic Modelling. Springer International Publishing, 2018

    2018

  20. [20]

    Probability Theory and Stochastic Modelling

    René Carmona and François Delarue.Probabilistic Theory of Mean Field Games with Applications II: Mean Field Games with Common Noise and Master Equations. Probability Theory and Stochastic Modelling. Springer International Publishing, 2018

    2018

  21. [21]

    Quantitative Convergence for Mean Field Control with Common Noise and Degenerate Idiosyncratic Noise.arXiv preprint arXiv:2409.14053, 2024

    Alekos Cecchin, Samuel Daudin, Joe Jackson, and Mattia Martini. Quantitative Convergence for Mean Field Control with Common Noise and Degenerate Idiosyncratic Noise.arXiv preprint arXiv:2409.14053, 2024

    work page arXiv 2024

  22. [22]

    Viscosity Solutions of a Class of Second-Order Hamilton–Jacobi–Bellman Equations in the Wasserstein Space.Appl

    Hang Cheung, Ho Man Tai, and Jinniao Qiu. Viscosity Solutions of a Class of Second-Order Hamilton–Jacobi–Bellman Equations in the Wasserstein Space.Appl. Math. Optim., 91(1):23, 2025

    2025

  23. [23]

    A Partial Laplacian as an Infinitesimal Generator on the Wasserstein Space.J

    Yat Tin Chow and Wilfrid Gangbo. A Partial Laplacian as an Infinitesimal Generator on the Wasserstein Space.J. Differential Equations, 267(10):6065–6117, 2019

    2019

  24. [24]

    Master Bell- man equation in the Wasserstein space: uniqueness of viscosity solutions.Trans

    Andrea Cosso, Fausto Gozzi, Idris Kharroubi, Huyên Pham, and Mauro Rosestolato. Master Bell- man equation in the Wasserstein space: uniqueness of viscosity solutions.Trans. Amer. Math. Soc., 377(1):31–83, 2024

    2024

  25. [25]

    Crandall, Hitoshi Ishii, and Pierre-Louis Lions

    Michael G. Crandall, Hitoshi Ishii, and Pierre-Louis Lions. User’s guide to viscosity solutions of second order partial differential equations.Bull. Amer. Math. Soc. (N.S.), 27(1):1–67, 1992

    1992

  26. [26]

    Well-posedness of Hamilton–Jacobi equations in the Wasserstein space: non-convex Hamiltonians and common noise.Communications in Partial Differential Equations, 50(1–2):1–52, 2025

    Samuel Daudin, Joe Jackson, and Benjamin Seeger. Well-posedness of Hamilton–Jacobi equations in the Wasserstein space: non-convex Hamiltonians and common noise.Communications in Partial Differential Equations, 50(1–2):1–52, 2025

    2025

  27. [27]

    A comparison principle for semilinear Hamilton–Jacobi– Bellman equations in the Wasserstein space.Calc

    Samuel Daudin and Benjamin Seeger. A comparison principle for semilinear Hamilton–Jacobi– Bellman equations in the Wasserstein space.Calc. Var. Partial Differential Equations, 63(4), 2024

    2024

  28. [28]

    Finite Dimensional Approximations of Hamilton–Jacobi–Bellman Equations in Spaces of Probability Measures.SIAM J

    Wilfrid Gangbo, Sergio Mayorga, and Andrzej Święch. Finite Dimensional Approximations of Hamilton–Jacobi–Bellman Equations in Spaces of Probability Measures.SIAM J. Math. Anal., 53(2):1320–1356, 2021. 27

    2021

  29. [29]

    Mészáros

    Wilfrid Gangbo and Alpár R. Mészáros. Global Well-Posedness of Master Equations for Determinis- tic Displacement Convex Potential Mean Field Games.Comm.PureAppl.Math., 75(12):2685–2801, 2022

    2022

  30. [30]

    Mészáros, Chenchen Mou, and Jianfeng Zhang

    Wilfrid Gangbo, Alpár R. Mészáros, Chenchen Mou, and Jianfeng Zhang. Mean Field Games Master Equations with Nonseparable Hamiltonians and Displacement Monotonicity.Ann. Probab., 50(6):2178–2217, 2022

    2022

  31. [31]

    Hamilton–Jacobi–Bellman Equations for the Optimal Control of the Duncan–Mortensen–Zakai Equation.J

    Fausto Gozzi and Andrzej Święch. Hamilton–Jacobi–Bellman Equations for the Optimal Control of the Duncan–Mortensen–Zakai Equation.J. Funct. Anal., 172(2):466–510, 2000

    2000

  32. [32]

    Springer, New York, third edition, 2014

    Loukas Grafakos.Modern Fourier Analysis, volume 250 ofGraduate Texts in Mathematics. Springer, New York, third edition, 2014

    2014

  33. [33]

    P.-L. Lions. Viscosity solutions of fully nonlinear second-order equations and optimal stochastic control in infinite dimensions. I. The case of bounded stochastic evolutions.Acta Math., 161:243– 278, 1988

    1988

  34. [34]

    P.-L. Lions. Lectures at collège de france. www.college-de-france.fr, 2011

    2011

  35. [35]

    Viscosity solutions of fully nonlinear second-order equations and optimal stochas- tic control in infinite dimensions

    Pierre-Louis Lions. Viscosity solutions of fully nonlinear second-order equations and optimal stochas- tic control in infinite dimensions. III. Uniqueness of viscosity solutions for general second-order equations.J. Funct. Anal., 86(1):1–18, 1989

    1989

  36. [36]

    Viscosity solutions of fully nonlinear second order equations and optimal stochas- tic control in infinite dimensions

    Pierre-Louis Lions. Viscosity solutions of fully nonlinear second order equations and optimal stochas- tic control in infinite dimensions. Part II: optimal control of Zakai’s equation. InStochastic Partial Differential Equations and Applications II, pages 147–170. Springer, 1989

    1989

  37. [37]

    Kolmogorov equations on spaces of measures associated to nonlinear filtering pro- cesses.Stochastic Process

    Mattia Martini. Kolmogorov equations on spaces of measures associated to nonlinear filtering pro- cesses.Stochastic Process. Appl., 161:385–423, 2023

    2023

  38. [38]

    Kolmogorov equations on the space of probability measures associated to the non- linear filtering equation: the viscosity approach.Stochastic Anal

    Mattia Martini. Kolmogorov equations on the space of probability measures associated to the non- linear filtering equation: the viscosity approach.Stochastic Anal. Appl., 42(6):987–999, 2024

    2024

  39. [39]

    Bellman equation and viscosity solutions for mean-field stochastic control problem.ESAIM Control Optim

    Huyên Pham and Xiaoli Wei. Bellman equation and viscosity solutions for mean-field stochastic control problem.ESAIM Control Optim. Calc. Var., 24(1):437–461, 2018

    2018

  40. [40]

    Mete Soner

    H. Mete Soner. On the Hamilton–Jacobi–Bellman equations in Banach spaces.J. Optim. Theory Appl., 57(3):429–437, 1988

    1988

  41. [41]

    Mete Soner and Qinxin Yan

    H. Mete Soner and Qinxin Yan. Viscosity Solutions for McKean–Vlasov Control on a Torus.SIAM J. Control Optim., 62(2):903–923, 2024

    2024

  42. [42]

    Mete Soner and Qinxin Yan

    H. Mete Soner and Qinxin Yan. Viscosity Solutions of the Eikonal Equation on the Wasserstein Space.Appl. Math. Optim., 90(1):Paper No. 1, 2024

    2024

  43. [43]

    A Finite-Dimensional Approximation for Partial Differential Equations on Wasserstein Space.Stochastic Process

    Mehdi Talbi. A Finite-Dimensional Approximation for Partial Differential Equations on Wasserstein Space.Stochastic Process. Appl., 177:104445, 2024

    2024

  44. [44]

    Viscosity Solutions for HJB Equations on the Process Space: Application to Mean Field Control with Common Noise.arXiv preprint arXiv:2401.04920, 2024

    Jianjun Zhou, Nizar Touzi, and Jianfeng Zhang. Viscosity Solutions for HJB Equations on the Process Space: Application to Mean Field Control with Common Noise.arXiv preprint arXiv:2401.04920, 2024. 28

    work page arXiv 2024