arxiv: 2605.11785 · v1 · submitted 2026-05-12 · 🧮 math.PR

Recognition: 2 theorem links

· Lean Theorem

Regularization of a mean-field SDE by an additive common noise: The conditional expectation case

Pierre Cardaliaguet (CEREMADE) , Benjamin Jourdain (CERMICS , MATHRISK)

Authors on Pith no claims yet

Pith reviewed 2026-05-13 05:15 UTC · model grok-4.3

classification 🧮 math.PR

keywords McKean-Vlasov SDEcommon noiseconditional expectationweak solutionspropagation of chaosmean-field interactionregularization by noise

0 comments

The pith

An additive common noise ensures existence and uniqueness for McKean-Vlasov SDEs with bounded measurable drifts depending on position and conditional expectation.

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

The paper establishes that adding both a common noise and an individual noise to a mean-field stochastic differential equation allows weak solutions to exist and be unique even when the drift function is merely bounded and measurable with respect to the particle's position and the conditional expectation of the positions. Without the individual noise, the same conclusion holds as long as the drift is also Lipschitz continuous in the position. A sympathetic reader would care because this result shows how a finite-dimensional common noise can regularize otherwise intractable discontinuities in the mean-field interaction, but only when the interaction is specifically through the conditional law rather than the unconditional distribution. This extends the applicability of mean-field models to settings with rougher, more realistic interaction rules.

Core claim

In the presence of an additive individual noise, existence and uniqueness of a weak solution hold for any drift given by a bounded and measurable function of the position and the conditional expectation. When there is no individual noise, existence and uniqueness still hold if the drift is in addition Lipschitz in the position variable. This demonstrates that the presence of a finite dimensional common noise may allow to overcome the discontinuity of the drift with respect to the interaction term, provided that this interaction term is a conditional expectation. The paper also proves propagation of chaos for systems of particles where the conditional expectation is replaced by the empirical

What carries the argument

The additive common noise acting together with interaction through the conditional expectation of positions, which provides regularization that overcomes discontinuities in the drift.

If this is right

Unique weak solutions exist for the SDE under the given conditions on the drift.
Propagation of chaos holds when approximating the conditional expectation by empirical means in particle systems.
The regularization works for drifts that are Lipschitz in position even without individual noise.
The approach extends to related particle systems with specially prepared noise.

Where Pith is reading between the lines

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

The conditional structure may allow similar noise-based regularization in other filtered or partial-information mean-field models.
Numerical simulations of the particle systems could test whether propagation of chaos persists for highly discontinuous drifts.
This mechanism might connect to filtering or control problems where conditional expectations arise naturally.

Load-bearing premise

The interaction must occur specifically through the conditional expectation rather than the unconditional law or empirical measure.

What would settle it

Constructing a bounded measurable drift depending on position and conditional expectation for which no weak solution exists despite the presence of both common and individual additive noises.

read the original abstract

We investigate a McKean-Vlasov stochastic differential equation with an additive common noise and in which the interaction is through the conditional expectation. We show that, in the presence of an additive individual noise, existence and uniqueness of a weak solution hold for any drift given by a bounded and measurable function of the position and the conditional expectation. When there is no individual noise, existence and uniqueness still hold if the drift is in addition Lipschitz in the position variable. This shows that the presence of a finite dimensional common noise may allow to overcome the discontinuity of the drift with respect to the interaction term, provided that this interaction term is a conditional expectation. We also prove propagation of chaos for systems of particles where the conditional expectation is replaced by the empirical mean of the positions or by a closely related contribution with better prepared noise.

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.

Desk Editor's Note private letter to a colleague

The paper shows that common noise regularizes bounded measurable drifts in McKean-Vlasov SDEs precisely when the interaction uses conditional expectation with respect to the common noise filtration.

read the letter

This result is new in its focus on the conditional-expectation interaction: with additive individual noise, weak existence and uniqueness hold for any bounded measurable drift depending on position and the conditional expectation; without individual noise, they add a Lipschitz condition in position. The common noise overcomes the discontinuity because the conditional structure lets the filtration absorb the jumps in a way that standard unconditional mean-field interactions do not. They also prove propagation of chaos for the corresponding particle systems, replacing the conditional expectation by the empirical mean or a related term with prepared noise. That part is useful and cleanly stated. The proofs appear to rest on standard weak-solution techniques for SDEs with common noise, adapted to the conditional setting, and the stress-test found no internal contradictions or circularity. The main limitation is the narrow scope: the regularization works only because the interaction is specifically the conditional expectation, not a general function of the law or empirical measure. If the drift depended on the unconditional law instead, the same argument would not go through. The paper does not claim broader applicability, so this is not a flaw but a boundary on the result. For readers working on mean-field limits, common-noise regularization, or non-Lipschitz drifts in stochastic analysis, the paper is worth reading and citing in that niche. It is precise enough and grounded enough to deserve peer review rather than a desk reject.

Referee Report

2 major / 3 minor

Summary. The paper studies a McKean-Vlasov SDE driven by an additive common noise in which the drift depends on the current position and the conditional expectation of the position with respect to the common-noise filtration. It proves existence and uniqueness of weak solutions when the drift is any bounded measurable function of these two arguments provided an individual Brownian motion is also present; when the individual noise is absent, the same conclusion holds if the drift is additionally Lipschitz continuous in the position variable. The manuscript further establishes a propagation-of-chaos result for the corresponding N-particle system in which the conditional expectation is replaced by the empirical mean (or a closely related term with suitably prepared noise).

Significance. If the stated results hold, the work isolates a precise structural condition—the use of conditional expectation with respect to the common noise—under which an additive finite-dimensional common noise regularizes otherwise discontinuous mean-field interactions. This supplies a new mechanism for obtaining well-posedness in McKean-Vlasov equations beyond the classical Lipschitz or monotonicity assumptions and directly supports the analysis of particle approximations in the presence of shared randomness.

major comments (2)

[§3.2, Theorem 3.4] §3.2, Theorem 3.4 (no-individual-noise case): the uniqueness argument invokes a Lipschitz condition in the position variable to control the difference of two conditional expectations, yet the estimate appears to rely on an implicit uniform integrability step that is not spelled out; without it the contraction may fail when the common noise is only finite-dimensional.
[§4.1, Proposition 4.2] §4.1, Proposition 4.2 (propagation of chaos): the convergence of the empirical-measure term to the conditional expectation is stated for bounded measurable drifts, but the proof sketch does not verify that the required tightness of the joint law of (position, conditional expectation) is preserved under the particle approximation when the drift is merely measurable.

minor comments (3)

[§2] Notation for the common-noise filtration is introduced in §2 but used without re-statement in later sections; a short reminder of the definition would improve readability.
[Abstract and §3.1] The abstract claims 'any bounded and measurable function' yet the statements of the main theorems include an implicit integrability condition on the drift; this minor mismatch should be aligned.
[Figure 1] Figure 1 (schematic of the common-noise filtration) is helpful but the caption does not indicate the dimension of the common noise; adding this detail would clarify the finite-dimensional assumption.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on our manuscript. We appreciate the positive assessment of the significance of the results. We address the two major comments point by point below.

read point-by-point responses

Referee: [§3.2, Theorem 3.4] §3.2, Theorem 3.4 (no-individual-noise case): the uniqueness argument invokes a Lipschitz condition in the position variable to control the difference of two conditional expectations, yet the estimate appears to rely on an implicit uniform integrability step that is not spelled out; without it the contraction may fail when the common noise is only finite-dimensional.

Authors: We thank the referee for this observation. In the uniqueness proof for Theorem 3.4, the Lipschitz condition in the position variable is applied to bound the difference of the drifts, leading to an estimate that involves the conditional expectation of the process difference. The required uniform integrability follows from the uniform second-moment bounds on the solutions, which are obtained via standard Gronwall arguments for SDEs with linear growth (guaranteed by the Lipschitz assumption). These bounds are independent of the specific solution and hold regardless of the dimension of the common noise. We agree, however, that this step is not spelled out explicitly enough and could lead to ambiguity. We will revise the manuscript to include a detailed paragraph making the uniform integrability argument fully explicit, using the moment estimates from the SDE. revision: yes
Referee: [§4.1, Proposition 4.2] §4.1, Proposition 4.2 (propagation of chaos): the convergence of the empirical-measure term to the conditional expectation is stated for bounded measurable drifts, but the proof sketch does not verify that the required tightness of the joint law of (position, conditional expectation) is preserved under the particle approximation when the drift is merely measurable.

Authors: We appreciate this comment. In Proposition 4.2 the boundedness of the drift guarantees uniform moment bounds for both the mean-field limit and the N-particle system, independent of N. These bounds yield tightness of the empirical measures in the space of probability measures on the path space. For the joint law of the position and the (empirical) conditional expectation, the common noise is shared across particles, so the conditional structure is preserved; tightness of the joint processes then follows from the uniform integrability induced by the bounded drift. We acknowledge that the current proof sketch is brief on this point. We will expand the argument in the revised manuscript to explicitly verify the preservation of tightness for the joint laws under the particle approximation. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained via standard analysis

full rationale

The paper proves existence and uniqueness of weak solutions for the McKean-Vlasov SDE under the stated conditions on the drift (bounded measurable or with added Lipschitz) and the interaction via conditional expectation w.r.t. common noise. These results rely on classical stochastic analysis tools (e.g., tightness, Skorokhod representation, fixed-point arguments for the conditional law) rather than any self-referential definitions, fitted parameters renamed as predictions, or load-bearing self-citations that reduce the central claim to its inputs. The key structural hypothesis—that interaction occurs specifically through the conditional expectation—is explicitly isolated as the feature enabling regularization of discontinuities and is not smuggled in via prior work by the authors; propagation of chaos is handled as an independent extension. No step equates a derived quantity to its own construction by definition.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on standard stochastic calculus axioms without new free parameters or invented entities.

axioms (1)

standard math Standard properties of Brownian motions, filtrations, and conditional expectations on a probability space
Implicit in the definition of the SDE and weak solutions

pith-pipeline@v0.9.0 · 5450 in / 1021 out tokens · 44361 ms · 2026-05-13T05:15:57.065144+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

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

IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

We show that, in the presence of an additive individual noise, existence and uniqueness of a weak solution hold for any drift given by a bounded and measurable function of the position and the conditional expectation.
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

the quantity Y = X − E[X|W0] formally satisfies a “regular” McKean-Vlasov equation

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

32 extracted references · 32 canonical work pages

[1]

(2016), Conce ntration inequalities: A Nonasymptotic Theory of Independence, Oxford

Boucheron, S., Lugosi, G., and Massart, P. (2016), Conce ntration inequalities: A Nonasymptotic Theory of Independence, Oxford

work page 2016
[2]

Cardaliaguet, P., Delarue, F., Lasry, J. M. and Lions, P. -L. (2019). The Master Equation and the Convergence Problem in Mean Field Games (AMS-201) (Vol. 381). Princeton University Press

work page 2019
[3]

Carlen, E., and Kree, P. (1991). Lp estimates on iterated stochastic integrals. The Annals of P roba- bility, 354-368

work page 1991
[4]

Carmona, R., and Delarue, F. (2018). Probabilistic theory of mean ﬁeld games with applications I -II. Berlin: Springer Nature

work page 2018
[5]

Carmona, R., Delarue, F., and Lacker, D. (2016). Mean ﬁel d games with common noise. Annals of probability: An oﬃcial journal of the Institute of Mathemat ical Statistics, 44(6), 3740-3803

work page 2016
[6]

Coghi, M., and Flandoli, F. (2016). Propagation of chaos for interacting particles subject to envi- ronmental noise. Annals of applied probability: an oﬃcial j ournal of the Institute of Mathematical Statistics, 26(3), 1407-1442

work page 2016
[7]

Coghi, M., and Gess, B. (2019). Stochastic nonlinear Fok ker-Planck equations. Nonlinear Analysis, 187, 259-278

work page 2019
[8]

Crowell, R. A. (2024). Emergence of regularity for limit points of McKean-Vlasov particle systems. arXiv preprint arXiv:2409.03459

work page arXiv 2024
[9]

Crowell, R. A. (2025). A new approach to stochastic McKea n-Vlasov limits with low-regularity coef- ﬁcients. arXiv preprint arXiv:2504.11228

work page arXiv 2025
[10]

Dawson, D., and Vaillancourt, J. (1995). Stochastic Mc Kean-Vlasov equations. Nonlinear Diﬀerential Equations and Applications NoDEA, 2(2), 199-229

work page 1995
[11]

and Tanr´ e, E

Delarue, F., Maillet, R. and Tanr´ e, E. (2024). Ergodic ity of some stochastic Fokker-Planck equations with additive common noise. arXiv preprint arXiv:2405.099 50

work page 2024
[12]

Delarue, F., and Mou, C. (2025). Major-minor mean ﬁeld g ames: common noise helps. arXiv preprint arXiv:2501.02627. 36 PIERRE CARDALIAGUET AND BENJAMIN JOURDAIN

work page arXiv 2025
[13]

Delarue, F., and Tchuendom, R. F. (2020). Selection of e quilibria in a linear quadratic mean-ﬁeld game. Stochastic Processes and their Applications, 130(2) , 1000-1040

work page 2020
[14]

Delarue, F., and Vasileiadis, A. (2025). Exploration n oise for learning linear-quadratic mean ﬁeld games. Mathematics of Operations Research, 50(3), 1762-18 31

work page 2025
[15]

Fournier, N., and Guillin, A. (2015). On the rate of conv ergence in Wasserstein distance of the empirical measure. Probability theory and related ﬁelds, 162(3), 707 -738

work page 2015
[16]

R., Siska, D., and Szpruch, L

Hammersley, W. R., Siska, D., and Szpruch, L. (2021). We ak existence and uniqueness for McKean- Vlasov SDEs with common noise. The Annals of Probability 202 1, Vol. 49, No. 2, 527-555

work page 2021
[17]

E., and Wang, Z

Jabin, P. E., and Wang, Z. (2016). Mean ﬁeld limit and pro pagation of chaos for Vlasov systems with bounded forces. Journal of Functional Analysis, 271(12), 3 588-3627

work page 2016
[18]

E., and Wang, Z

Jabin, P. E., and Wang, Z. (2018). Quantitative estimat es of propagation of chaos for stochastic systems with W 1−∞ kernels. Inventiones mathematicae, 214(1), 523-591

work page 2018
[19]

Jabir, J. F. (2019). Rate of propagation of chaos for diﬀ usive stochastic particle systems via Girsanov transformation. arXiv preprint arXiv:1907.09096

work page arXiv 2019
[20]

Jourdain, B. (1997). Diﬀusions with a nonlinear irregu lar drift coeﬃcient and probabilistic interpre- tation of generalized Burgers’ equations. ESAIM: Probabil ity and Statistics, 1, 339-355

work page 1997
[21]

Lacker, D., (2018) On a strong form of propagation of cha os for McKean-Vlasov equations, Electron. Commun. Probab., Volume 23, paper no. 45, 11

work page 2018
[22]

Lacker, D., Shkolnikov, M., and Zhang, J. (2022). Super position and mimicking theorems for condi- tional McKean-Vlasov equations. Journal of the European Ma thematical Society, 25(8), 3229-3288

work page 2022
[23]

Karatzas, I., and Shreve, S. (2014). Brownian motion an d stochastic calculus. springer

work page 2014
[24]

G., and Xiong, J

Kurtz, T. G., and Xiong, J. (1999). Particle representa tions for a class of nonlinear SPDEs. Stochastic Processes and their Applications, 83(1), 103-126

work page 1999
[25]

Maillet, R. (2023). A note on the Long-Time behaviour of Stochastic McKean-Vlasov Equations with common noise. arXiv preprint arXiv:2306.16130

work page arXiv 2023
[26]

Nikolaev, P. (2025). Fluctuation for interacting part icle systems with common noise. arXiv preprint arXiv:2504.05044

work page arXiv 2025
[27]

Sznitman, A. S. (2006). Topics in propagation of chaos. In Ecole d’´ et´ e de probabilit´ es de Saint-Flour XIX—1989 (pp. 165-251). Berlin, Heidelberg: Springer Berl in Heidelberg

work page 2006
[28]

Mishura, Y., and Veretennikov, A. (2020). Existence an d uniqueness theorems for solutions of McKean- Vlasov stochastic equations. Theory of Probability and Mat hematical Statistics, 103, 59-101

work page 2020
[29]

Shkolnikov, M., and Yeung, L. C. (2026). From rank-base d models with common noise to pathwise entropy solutions of SPDEs. The Annals of Applied Probabili ty, 36(1), 319-354

work page 2026
[30]

Sznitman, A. S. (2006). Topics in propagation of chaos. In Ecole d’´ et´ e de probabilit´ es de Saint-Flour XIX, 1989 (pp. 165-251). Berlin, Heidelberg: Springer Berl in Heidelberg

work page 2006
[31]

Vaillancourt, J. (1988). On the existence of random McK ean-Vlasov limits for triangular arrays of exchangeable diﬀusions. Stochastic Analysis and Applicat ions, 6(4), 431-446

work page 1988
[32]

Veretennikov, A. J. (1981). On strong solutions and exp licit formulas for solutions of stochastic integral equations. Mathematics of the USSR-Sbornik, 39(3), 387

work page 1981