Pith. sign in

REVIEW 2 major objections 4 minor 4 cited by

Conditional GMC within the stochastic heat flow

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Taking the random polymer measure of the stochastic heat flow at parameter $\theta$ as reference and applying an independent path-space Gaussian multiplicative chaos of noise strength $a$ produces, in law, the polymer measure at parameter…

desk verdict A genuinely new conditional GMC result for the critical 2D polymer measure, but the proof has a load-bearing gap in the dyadic uniqueness step; right verdict is conditional and it deserves a serious referee. read the letter →

arxiv 2507.16056 v1 pith:IZIQMMSE submitted 2025-07-21 math.PR

Jeremy Clark , Li-Cheng Tsai This is my paper · ORCID
classification math.PR MSC 60K3560H1560G57
keywords stochasticheatflowpolymermeasureGaussianmultiplicativechaosconditionalGMCdelta-Bosesemigroupintersectionlocaltimecriticaldimension1+2
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper proves that the polymer measures of the stochastic heat flow, indexed by a real disorder parameter $\theta$, are tied together by a conditional Gaussian multiplicative chaos (GMC) structure: exponentiating an independent Gaussian field against $M^{\theta}_{[s,t]}$ with noise strength $a$ yields a random measure equal in law to $M^{\theta+a}_{[s,t]}$. This matters because it makes the disorder parameter act like a time variable: the family $(M^{\theta})_{\theta\in\mathbb{R}}$ can be coupled so that $\theta$ drives a Markovian martingale, and the coupling turns into a tool for extracting concrete properties of the measures. The proof matches all annealed moments of the conditional GMC to the delta-Bose semigroup, then uses an axiomatic characterization to upgrade the moment match into an equality of laws. The main applications are that polymer measures and the stochastic heat flow tested against nonnegative functions are almost surely strictly positive, and that the stochastic heat flow converges to zero as $\theta\to\infty$.

What carries the argument

The load-bearing object is the intersection-local-time operator $T^{\theta,[s,t]}_{[s,t]}$ on $L^2(\Gamma_{[s,t]}, \overline{M}^{\theta}_{[s,t]})$, where $\overline{M}^{\theta}$ is the polymer measure with a fixed exponential weight: a positive Hilbert-Schmidt operator whose kernel records how much two polymer paths intersect. A Gaussian noise is attached to $M^{\theta}$ by factorizing this operator as $YY^*=T$; the standard factorization uses the eigenfunctions of $T$, and the GMC itself is defined through Kahane's martingale approximation. The proof then runs through three linked mechanisms: an operator computation of all annealed moments identifying them with the delta-Bose semigroup at parameter $\theta+a$, an operator-embedding result that lets eigenfunctions of the intersection-local-time operator be transported between nested time intervals even when the underlying measures may be mutually singular, and a universal-Hilbert-space coupling that makes the conditional GMCs on different intervals consistent. This coupling is what allows the axiomatic characterization to be applied and yields the GMC coupling in which $M^{\theta}$ is Markovian in $\theta$.

What would settle it

Compute the fifth annealed moment of the conditional GMC, $\mathbb{E}\mathbb{E}(G\circ|_{\{s,t\}}^{-1})^{\otimes 5}$, and compare it with $dx\,dx'\, Q^{[\![5]\!],\theta+a}(t-s,x,x')$; any mismatch for some $\theta,a,s<t$ would break the moment matching that Proposition 3.1 claims for all $n$, and would falsify the law identification in Theorem 1.1.

Watch

Extended reading notes

Core claim

The central claim, Theorem 1.1, is that for $s<t$, $\theta\in\mathbb{R}$, and $E\xi_1^2=a$, the conditional GMC $G[M^{\theta}_{[s,t]}, Y^{\theta}_{\star[s,t]}, \xi]$ is equal in law to $M^{\theta+a}_{[s,t]}$. Here $M^{\theta}_{[s,t]}$ is the polymer measure of the stochastic heat flow, $\xi$ is an independent iid Gaussian sequence with variance $a$, and $Y^{\theta}_{\star[s,t]}$ is the operator that factorizes the intersection-local-time kernel, the standard choice built from the eigenfunctions of that kernel. The authors emphasize that $M^{\theta}$ cannot be a GMC over a deterministic reference measure in dimension $1+2$, because paths are expected to localize on a fractal set; the conditional formulation replaces the deterministic reference by $M^{\theta_0}$ at a finite parameter below $\theta$, so that only a finite amount of noise is needed. Theorem 1.1 is established by proving the moment matching for all $n\in\mathbb{N}$ and then invoking the axiomatic characterization that turns matching the first four moments into equality in law.

Load-bearing premise

The argument relies on the axiomatic characterization that a random measure with independent increments on dyadic intervals, Chapman-Kolmogorov convolution, and first four moments matching the delta-Bose semigroup must have the same law as the polymer measure; if identifying the law required matching more than four moments, or if the dyadic restriction failed, the equality of laws in Theorem 1.1 would not follow.

Editorial extensions

If this is right

  • For every fixed $\theta_0$, the construction couples $M^{\theta_0}$ and $M^{\theta_0+a}$, and letting the Gaussian variables be Brownian motions in $a$ produces a joint coupling of all $M^{\theta}$ in which the family is Markovian and martingale in $\theta$.
  • For any deterministic nonnegative test function $f$ on path space that is not almost surely zero under the Wiener measure, $M^{\theta}_{[s,t]}f>0$ almost surely; consequently the partition function $Z^{\theta}_{s,t}g\otimes g'$ is almost surely positive, so normalized polymer measures are well defined.
  • The stochastic heat flow $Z^{\theta}_{s,t}$ converges to zero vaguely almost surely as $\theta\to\infty$, and a stronger version holds against single test functions satisfying a finite moment condition.
  • The equality $\mathbb{E}M^{\theta}_{[s,t]}=\text{Wein}_{[s,t]}$ and the fact that $M^{-\infty}$ is the Wiener measure explain why no classical GMC representation of $M^{\theta}$ exists: moving from the deterministic reference to parameter $\theta$ would require infinite noise strength, whereas moving between finite parameters requires only finite strength $\theta-\theta_0$.
  • The moment matching holds for every $n\in\mathbb{N}$, not just the first two moments, so the identification with the delta-Bose semigroup is not limited by the martingale techniques that previously gave only second moments.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the GMC coupling suggests one can perform stochastic-calculus-style arguments in the disorder parameter $\theta$; martingale estimates or Itô-type computations along $\theta$ could give quantitative rates for the vanishing in Corollary 1.4 rather than only almost sure convergence.
  • Editorial extension: the path-space construction bypasses spacetime white noise entirely, using only the intersection-local-time kernel; this is a template that could define conditional chaos on other singular random reference measures whose paths live on null sets.
  • Editorial extension: the proof identifies laws using only the first four moments through Proposition 1.5; a natural stress test is whether the same four-moment characterization holds for non-dyadic time grids, since the dyadic restriction is an artifact of the proof strategy.
  • Editorial extension: if the conjectured mutual singularity of the polymer measure and its time-$[s,t]$ marginal holds, the operator embedding result in the paper is the mechanism that makes the singularity harmless; verifying that singularity directly would test how essential this part of the construction is.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. The paper claims that the family of critical 2D polymer measures M^theta_{[s,t]} has a conditional Gaussian multiplicative chaos structure: for s<t, theta in R, and E(xi_1^2)=a, the GMC built with reference M^theta_{[s,t]} and an M^theta-measurable operator Y^theta_star is equal in law to M^{theta+a}_{[s,t]}. The proof combines: (i) an all-n moment matching between the conditional GMC and the delta-Bose semigroup (Proposition 3.1), (ii) an operator embedding result comparing intersection-local-time operators on nested path spaces (Proposition 4.1), (iii) a coupling of the Gaussian vectors over dyadic intervals (Proposition 5.1), and (iv) a dyadic axiomatic characterization of the polymer measure (Proposition 1.5) that converts moment equality into equality in law. Two applications are derived: a.s. strict positivity of M^theta against non-Weiner test functions (Corollary 1.3) and vague convergence of the SHF to 0 as theta tends to infinity (Corollary 1.4 and Corollary 1.4').

Significance. If the proof is completed, the main theorem is a substantial and genuinely new structural result: it gives an interlocking GMC picture for the whole family M^theta, a Markovian martingale coupling in theta, and new positivity and convergence results for the critical 2D stochastic heat flow. The moment matching for all n is a strong technical contribution, as is the operator-embedding mechanism that bypasses the expected mutual singularity of polymer measures on different time intervals. The applications in Corollaries 1.3 and 1.4 are concrete and falsifiable, and the result goes beyond the 1+1-dimensional GMC constructions. However, the manuscript as written contains two load-bearing deferred steps: the dyadic characterization in Proposition 1.5 and the operator identity (6.6) supporting Proposition 1.2. Both need to be repaired or made fully explicit before the central claims are established.

major comments (2)
  1. [Appendix A / Proposition 1.5] Proposition 1.5 is the hinge that converts the moment identity of Proposition 3.1 into the equality in law of Theorem 1.1, but its proof is only three sentences and does not bridge the gap between the dyadic hypotheses and the full-time theorem [Tsa24, Theorem 1.9] recalled in Section 1.2. The recalled theorem characterizes a continuous M_+(R^4)-valued process indexed by all s<t, whereas Proposition 1.5 assumes only s<t in qD and verifies Conditions (2)-(4) on qD. The appendix neither constructs an extension of Z'_{s,t} to all real times nor verifies continuity and Axiom (1) for that extension. If [Tsa24, Theorem 1.9] has no dyadic version, the conclusion of Proposition 1.5 does not follow; if a dyadic version exists, it should be stated and proved or explicitly quoted. This is load-bearing because without Proposition 1.5, Proposition 3.1 gives only moment equality and not equality in law.
  2. [Section 6.1, Eq. (6.6)] The proof of Proposition 1.2 hinges on the identity Y^phi_star Y^phi*_star = T^{theta0} in L^2(Gamma, M^{theta0}), stated as (6.6). The text says this follows by applying Lemmas 4.4 and 4.7 with (mu, mu_l, mu', k) = (M^{theta0}, M'_ell, M^{theta1}, tau), and that the verification of Assumptions 4.3 and 4.6 is 'similar' to earlier arguments, 'which we do not repeat.' This is not a routine repetition: the sequence M'_ell is itself defined through a GMC martingale approximation, and conditions (4.16)-(4.20) have to be checked for this specific sequence. Since Proposition 1.2 underlies both Corollaries 1.3 and 1.4, the deferred verification should be supplied or at least the reduction to the earlier proofs made explicit.
minor comments (4)
  1. [Section 1.2] The symbol q is overloaded: it is used both as the right endpoint in the renamed statement of Theorem 1.1 and as the base of the dyadic grid qD. This makes sentences such as 'take r=0' and 'for q in (0,infinity)' hard to parse; consider renaming the endpoint.
  2. [Section 2.3, Eq. (2.33)] The notation G o |_{s,t}^{-1} is used before pushforwards by evaluation maps are formally introduced; please define this operation at first use.
  3. [Section 6.1] Even if the verification of (6.6) is accepted as a repetition of earlier arguments, it would be helpful to state explicitly which of Conditions (4.16)-(4.20) are being verified and where. This would also clarify the role of the exponential weighting in M'_ell.
  4. [References] The manuscript relies heavily on [Tsa24, Theorem 1.9] and [CM24, Proposition 2.12]; since [Tsa24] is a preprint, please quote the exact statement of Theorem 1.9 in the revision so that the dyadic characterization can be checked without access to that preprint.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: Theorem 1.1 follows from a genuine moment computation against the externally defined delta-Bose semigroup, with the law identification supplied by an external uniqueness theorem that does not assume the conclusion.

full rationale

I find no circular step in which a claimed prediction reduces by construction to its own input. Theorem 1.1 identifies the law of the conditional GMC G[M^theta, Y^theta, xi] with the polymer measure M^{theta+a}. The target object M^{theta+a} is defined externally through the SHF polymer measure of [CM24] and the delta-Bose moment semigroup Q^{theta+a} of [GQT21]; it is not defined in terms of the GMC. The moment matching in Proposition 3.1 is a substantive computation: the annealed expectation (3.19) is expanded in a series, and the resummation in (3.37) proves E^{theta,a}(t) = Q^{theta+a}(t). This is a diagrammatic calculation, not a definitional identity. The conversion from moments to equality in law uses Proposition 1.5, whose proof in Appendix A invokes [Tsa24, Theorem 1.9]. This is indeed a self-citation, since Tsai is an author, and it is load-bearing for the law identification. However, the cited theorem is a separate axiomatic characterization of the SHF whose stated assumptions do not include the conditional GMC or Theorem 1.1, so under the hard rule on independent support it is real evidence rather than circularity. The weakest point is the dyadic restriction in Proposition 1.5: Appendix A asserts that Conditions (2)-(4) on q-dyadic times plus [Tsa24, Theorem 1.9] identify the law, but it does not explicitly construct a full-time continuous extension of Z' and verify Axiom (1). That is an omitted justification and a correctness risk, not a circular reduction, because the cited theorem is not equivalent to the conclusion. There are no fitted parameters, no statistically forced predictions, and no known result is merely renamed. Accordingly the circularity score is 0.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The paper introduces no fitted constants. It rests on prior constructions of the SHF and polymer measure, the delta-Bose semigroup, intersection local times, and standard Gaussian and martingale tools. These are external results rather than assumptions tailor-made to prove Theorem 1.1.

assumptions (6)
  • domain assumption The SHF is unique and characterized by the moment axioms with moments n=1,2,3,4 given by the delta-Bose semigroup, from [Tsa24, Theorem 1.9].
    Invoked in Proposition 1.5 and Section 1.2 to turn moment matching into equality in law for the conditional GMC.
  • domain assumption The polymer measure M^theta_[s,t] exists as a locally finite random measure on continuous path space with finite-dimensional marginals given by SHF convolutions, from [CM24].
    Used throughout as the reference measure in Section 2.1.
  • standard math The delta-Bose semigroup Q^[n],theta defined by the series (3.9) converges and satisfies operator bounds (3.11), from [GQT21].
    Used in the moment matching and resummation arguments in Section 3.1.
  • domain assumption The intersection local time tau exists with no atoms and satisfies convergence (2.12), and the operator T is positive, from [CM23] and Lemma A.3.
    Needed to define the operator factorization Y^theta and the conditional GMC in Section 2.2.
  • domain assumption The martingale consistency relation and sigma-algebra union in Proposition 2.1 hold.
    Used to compare L^2 spaces under potentially mutually singular measures in Section 4 and to couple GMCs over different time intervals in Section 5.
  • standard math The Gaussian zero-one law from [Bog98, Theorem 2.5.2] holds.
    Used in Lemma 6.1 to prove the positivity comparison principle.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Conditional GMC within the stochastic heat flow." pith.science (2026). https://pith.science/paper/IZIQMMSE

@misc{pith2026250716056,
  author       = {Pith},
  title        = {Pith review of: Conditional GMC within the stochastic heat flow},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IZIQMMSE}},
  note         = {Machine review of arXiv:2507.16056}
}
abstract

We establish that the family of polymer measures $M^{\theta}_{[s,t]}$ associated with the Stochastic Heat Flow (SHF), indexed by $\theta\in\mathbb{R}$, has a conditional Gaussian Multiplicative Chaos (GMC) structure. Namely, taking the random measure $M^{\theta}_{[s,t]}$ as the reference measure, we construct the path-space GMC with noise strength $a > 0$ and prove that the resulting random measure is equal in law to $M^{\theta+a}_{[s,t]}$. As two applications, we prove that the polymer measure and SHF tested against general nonnegative functions are almost surely strictly positive and that the SHF converges to $0$ as $\theta\to\infty$.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Sharp moment and upper tail asymptotics for the critical $2d$ Stochastic Heat Flow

    math.PR 2025-07 conditional novelty 8.0 of 10

    The h-th moment of the critical 2d Stochastic Heat Flow mass is at least exp(exp(c h)), matching a 1999 prediction and exponentially improving the known lower bound.

  2. Extrema of cooling branching Brownian motion and related Gaussian fields

    math.PR 2026-06 unverdicted novelty 7.0 of 10

    Introduces slowed-down Gaussian fields (including 1D branching Brownian motions in cooling environments) and proves tightness of maxima with growth T^{1-α} and phase transition at α=1/3.

  3. McKean-Vlasov limits of scaling-critical reaction-diffusion equations with random initial data

    math.PR 2025-09 conditional novelty 7.0 of 10

    For a class of scaling-critical reaction-diffusion equations with log-attenuated nonlinearities and white-noise initial data, the solutions converge to a Gaussian McKean-Vlasov process whose variance is set by an ODE.

  4. Enhanced noise sensitivity, 2D directed polymers and Stochastic Heat Flow

    math.PR 2025-07 conditional novelty 7.0 of 10

    A general, rate-optimal BKS noise-sensitivity criterion is proven, and it yields the independence of the critical 2D Stochastic Heat Flow from the disorder white noise.

Reference graph

Works this paper leans on

25 extracted references · 19 canonical work pages · cited by 4 Pith papers

  1. [1]

    The intermediate disorder regime for directed polymers in dimension 1+1

    Tom Alberts, Konstantin Khanin, and Jeremy Quastel. The intermediate disorder regime for directed polymers in dimension 1+1. Ann Probab , 42(3):1212--1256, 2014

    work page 2014

  2. [2]

    Handbook of mathematical functions: with formulas, graphs, and mathematical tables , volume 55

    Milton Abramowitz and Irene A Stegun. Handbook of mathematical functions: with formulas, graphs, and mathematical tables , volume 55. Courier Corporation, 1965

    1965

  3. [3]

    Bogachev

    Vladimir I. Bogachev. Gaussian measures , volume 62 of Mathematical Surveys and Monographs . American Mathematical Society, Providence, RI, 1998

    work page 1998

  4. [4]

    Continuum models of directed polymers on disordered diamond fractals in the critical case

    Jeremy Clark. Continuum models of directed polymers on disordered diamond fractals in the critical case. Ann Appl Probab , 32(6):4186--4250, 2022

    work page 2022

  5. [5]

    The conditional G aussian multiplicative chaos structure underlying a critical continuum random polymer model on a diamond fractal

    Jeremy Clark. The conditional G aussian multiplicative chaos structure underlying a critical continuum random polymer model on a diamond fractal. In Ann Inst H Poincar\' e Probab Statist , volume 59, pages 1203--1222. Institut Henri Poincar \'e , 2023

    work page 2023

  6. [6]

    On planar Brownian motion singularly tilted through a point potential

    Jeremy Clark and Barkat Mian. On planar B rownian motion singularly tilted through a point potential. arXiv:2306.14849 , 2023

    work page Pith review arXiv 2023

  7. [7]

    Continuum polymer measures corresponding to the critical 2d stochastic heat flow

    Jeremy Clark and Barkat Mian. Continuum polymer measures corresponding to the critical 2d stochastic heat flow. arXiv:2409.01510 , 2024

    arXiv 2024

  8. [8]

    The critical 2d stochastic heat flow

    Francesco Caravenna, Rongfeng Sun, and Nikos Zygouras. The critical 2d stochastic heat flow. Invent Math , 233(1):325--460, 2023

    work page 2023

Show all 25 references
  1. [9]

    The critical 2d stochastic heat flow is not a gaussian multiplicative chaos

    Francesco Caravenna, Rongfeng Sun, and Nikos Zygouras. The critical 2d stochastic heat flow is not a gaussian multiplicative chaos. Ann Probab , 51(6):2265--2300, 2023

    2023

  2. [10]

    The critical 2d stochastic heat flow and related models

    Francesco Caravenna, Rongfeng Sun, and Nikos Zygouras. The critical 2d stochastic heat flow and related models. arXiv:2412.10311 , 2024

    2024 arXiv

  3. [11]

    Singularity and regularity of the critical 2d stochastic heat flow

    Francesco Caravenna, Rongfeng Sun, and Nikos Zygouras. Singularity and regularity of the critical 2d stochastic heat flow. arXiv:2504.06128 , 2025

    2025 arXiv

  4. [12]

    H amiltonians for systems of N particles interacting through point interactions

    Gianfausto Dell'Antonio, Rodolfo Figari, and Alessandro Teta. H amiltonians for systems of N particles interacting through point interactions. In Annales de l'IHP Physique th \'e orique , volume 60, pages 253--290, 1994

    1994

  5. [13]

    Multi-particle S chr \"o dinger operators with point interactions in the plane

    Jonathan Dimock and Sarada Rajeev. Multi-particle S chr \"o dinger operators with point interactions in the plane. J Phys A: Math Gen , 37(39):9157, 2004

    2004

  6. [14]

    Moments of the 2d SHE at criticality

    Yu Gu, Jeremy Quastel, and Li-Cheng Tsai. Moments of the 2d SHE at criticality. Probab Math Phys , 2(1):179--219, 2021

    2021

  7. [15]

    Sur le chaos multiplicatif

    Jean-Pierre Kahane. Sur le chaos multiplicatif. Ann. Sci. Math. Qu\'ebec , 9(2):105--150, 1985

    1985

  8. [16]

    Foundations of modern probability , volume 99 of Probability Theory and Stochastic Modelling

    Olav Kallenberg. Foundations of modern probability , volume 99 of Probability Theory and Stochastic Modelling . Springer, Cham, third edition, 2021

    2021

  9. [17]

    Moreno Flores

    Gregorio R. Moreno Flores. On the (strict) positivity of solutions of the stochastic heat equation. Ann Probab , 42(4):1635--1643, 2014

    2014

  10. [18]

    On the support of solutions to the heat equation with noise

    Carl Mueller. On the support of solutions to the heat equation with noise. Stochastics Stochastics Rep , 37(4):225--245, 1991

    1991

  11. [19]

    KPZ fluctuations in the planar stochastic heat equation

    Jeremy Quastel, Alejandro Ramirez, and Balint Virag. KPZ fluctuations in the planar stochastic heat equation. arXiv:2210.13607 , 2022

    2022 arXiv

  12. [20]

    A condensation of interacting B osons in two dimensional space

    Sarada Rajeev. A condensation of interacting B osons in two dimensional space. arXiv: hep-th/9905120 , 1999

    1999 arXiv

  13. [21]

    Methods of modern mathematical physics

    Michael Reed and Barry Simon. Methods of modern mathematical physics. I . F unctional analysis . Academic Press, New York-London, 1972

    1972

  14. [22]

    Gaussian multiplicative chaos and applications: a review

    R\'emi Rhodes and Vincent Vargas. Gaussian multiplicative chaos and applications: a review. Probab Surv , 11:315--392, 2014

    2014

  15. [23]

    On G aussian multiplicative chaos

    Alexander Shamov. On G aussian multiplicative chaos. J Funct Anal , 270(9):3224--3261, 2016

    2016

  16. [24]

    Two dimensional delta- B ose gas in a weighted space

    Sudheesh Surendranath and Li-Cheng Tsai. Two dimensional delta- B ose gas in a weighted space. Electron Commun Probab , 30:1--10, 2025

    2025

  17. [25]

    Stochastic heat flow by moments

    Li-Cheng Tsai. Stochastic heat flow by moments. arXiv: 2410.14657 , 2024

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.