Pith. sign in

REVIEW 3 major objections 4 minor 69 references

Parallel spin wave for the Villain model

T0 review · 3 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 In dimension d=3, the low-temperature Villain model's truncated cos-cos correlation decays like |x|^{-2} up to a logarithmic factor, and in every d≥3 the upper bound improves the classical |x|^{2-d} rate; consequently the spin-wave…

desk verdict A real step on the Villain parallel correlation: d=3 gives the conjectured |x|^-2 up to log^22, but the d=3 upper bound rests on two unproved analytic inputs (B.5 and 4.11) that a referee must see. read the letter →

arxiv 2507.04098 v1 pith:6LZCYY6Q submitted 2025-07-05 math.PR math-phmath.MP

Paul Dario , Wei Wu This is my paper · ORCID
classification math.PRmath-phmath.MP MSC 60K3582B2082B2735B65
keywords Villainmodelspin-waveconjectureparallelcorrelationfunctionsHelffer–SjöstrandrepresentationrandominterfacemodelsellipticandparabolicregularityGoldstonebosonsXY
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 conjectured spin-wave decay for the parallel (cos-cos) correlation in the low-temperature Villain model is correct in dimension 3 up to a logarithmic correction, and it improves the known upper bound in every dimension d≥3. In d=3 the truncated correlation is shown to lie between constants times $|x|^{-2}$ and $(\ln|x|)^{22}|x|^{-2}$; in general d it lies between constants times $|x|^{-2(d-2)}$ and $(\ln|x|)^{4d+10}|x|^{1-d}$. Since the lower bound is already the spin-wave prediction, the result puts the massless fluctuation picture for this Abelian spin system on the right scale in all d≥3, with only a logarithmic gap in d=3. An immediate consequence is that the leading coefficients in the spin-spin and transversal correlation asymptotics are equal. A reader should care because this is the last piece of the Gaussian spin-wave prediction for the Villain model's basic two-point functions.

What carries the argument

The argument starts from the Fröhlich–Spencer duality that rewrites cosine-cosine expectations as observables of a vector-valued $\nabla\varphi$ interface model at large inverse temperature. The Helffer–Sjöstrand representation then turns covariances of those observables into solutions of an infinite-dimensional linear elliptic system, and the decay estimates are controlled by the associated Green's matrix and its codifferentials. For the leading cancellations the paper uses the second-order Helffer–Sjöstrand equation, which governs the derivative of the Green's matrix. Because the spatial operator has small ellipticity contrast of order $\beta^{-1}$, Schauder and Calderón–Zygmund arguments yield annealed $L^p$ regularity of the Langevin heat kernel; the logarithmic factors in the upper bound come from the infinite-range part of the operator. The lower bound uses instead the Dunlop–Newman correlation inequality for the Villain model, obtained via a metric-graph limit from the XY model, together with the earlier transversal asymptotics.

What would settle it

Run the Langevin dynamics for the interface model on a large periodic box at $\beta=50$ and compute the annealed norm $\lVert d_1^*d_2^*P(t,x,y)\rVert_{L^8}$; if it grows faster than $(\ln t)^{d+2}t^{-d/2-1}\exp(-|x-y|/(C\sqrt{t}))$ at large $t$, then the key estimates (4.4)–(4.5) fail and the upper bound's main input is gone. Independently, a Monte Carlo measurement of the d=3 parallel correlation at separations $|x|$ between 50 and 500 that falls faster than $1/|x|^2$ would contradict the lower bound.

Watch

Extended reading notes

Core claim

Theorem 2: for $d=3$ there is a critical inverse temperature $\beta_0$ and an exponent $\kappa=22$ such that for every $\beta\geq\beta_0$ and every $x\neq 0$, the truncated parallel correlation satisfies $C_2|x|^{-2} \leq \mathrm{cov}_{\mu_{\mathrm{Vil},\beta}}[\cos\theta(0),\cos\theta(x)] \leq C_1(\ln|x|_+)^{22}|x|^{-2}$. For general $d\geq3$ the same correlation satisfies $C_2|x|^{-2(d-2)} \leq \mathrm{cov} \leq C_1(\ln|x|_+)^{4d+10}|x|^{1-d}$. The upper bound improves the 1981 bound $|x|^{2-d}$ for the related XY model, and the lower bound ties the decay to the predicted $|x|^{-2(d-2)}$ free-field rate. The paper also draws the corollary that the constants $c_1$ and $c_2$ of the previously established spin-spin and transversal asymptotics agree.

Load-bearing premise

The whole d=3 upper bound leans on a technical heat-kernel decay estimate whose proof in the appendix is not written out: it cites a Calderón–Zygmund lemma, and the second-order Green's matrix estimate is stated with proof omitted.

Editorial extensions

If this is right

  • In dimension 3, for every sufficiently large $\beta$, the truncated cos-cos correlation lies within a logarithm of the free-field rate $|x|^{-2}$.
  • In every dimension $d\geq3$, the upper bound $(\ln|x|)^\kappa|x|^{1-d}$ improves the earlier $|x|^{2-d}$ upper bound for parallel correlations in the XY-type model.
  • The leading constants in the spin-spin and transversal correlation asymptotics coincide ($c_1=c_2$), fixing the coefficient of the Goldstone-mode contribution.
  • The lower bound $|x|^{-2(d-2)}$ holds in all $d\geq3$ at low temperature, matching the Gaussian prediction.
  • The constants in the correlation bounds scale like $1/\beta^2$, so the estimate sharpens as the temperature decreases.

Reading between the lines

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

  • The paper leaves open a log-free upper bound: the logarithmic factor appears only through the annealed codifferential estimates, so sharpening Appendix B would plausibly yield $\kappa=0$ in $d=3$.
  • The same two-step mechanism—Abelian duality to a $\nabla\varphi$ interface, then second-order Helffer–Sjöstrand regularity—should transfer to other lattice spin systems whose Villain action is a small perturbation of the Laplacian at low temperature.
  • A numerical evaluation of the Langevin heat-kernel estimates on finite tori would separate a genuine physical logarithmic correction from an artifact of the proof technique.
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

3 major / 4 minor

Summary. The paper studies the low-temperature Villain model on Z^d, d ≥ 3, and proves bounds for the truncated parallel correlation C(x) = ⟨cos θ(0) cos θ(x)⟩ − ⟨cos θ(0)⟩^2. Theorem 2 states that for d = 3, C2/|x|^2 ≤ C(x) ≤ C1 (ln|x|)^22 / |x|^2, and for general d ≥ 3, C2/|x|^{2(d−2)} ≤ C(x) ≤ C1 (ln|x|)^{4d+10} / |x|^{d−1}. This verifies the conjectured |x|^{−2(d−2)} decay in d = 3 up to a logarithmic factor, improves the BFL+81 upper bound |x|^{2−d} in general d, and yields Corollary 1.1 (c1 = c2 in the asymptotics of the authors' prior work [DW24]). The proof maps the Villain model via the Fröhlich–Spencer duality to a vector-valued ∇φ interface model, applies the Helffer–Sjöstrand representation, and then uses annealed elliptic and parabolic regularity for systems with small ellipticity contrast. The main chain — the duality identities of Proposition 3.1, the reduction to (3.12)–(3.13), the two applications of the Helffer–Sjöstrand formula, and the second-order equation estimates (3.25)–(3.27) — is written out in detail, and the logarithmic bookkeeping is consistent with κ = 4d + 10.

Significance. If correct, the d = 3 upper bound confirms the spin-wave prediction for the parallel correlation up to a logarithmic factor, the general-d bound improves the 1981 BFL+81 exponent by a full power of |x|, and Corollary 1.1 identifies the leading constants c1 = c2 in [DW24]. The proof strategy is natural and the core algebraic chain (duality, Helffer–Sjöstrand representation, convolution bounds of Appendix C) is presented at a level of detail that allows the arithmetic to be checked; the value κ = 4d + 10 is explicit and no free parameters are fitted. The paper deserves credit for being transparent about its dependencies: it explicitly flags in Remark B.6 that the parabolic-discrete Calderón–Zygmund criterion is an unproved adaptation of [AKM19, Lemma 7.2], in Proposition 4.11 that the second-order Green's matrix bounds are stated without proof, and in Remark 3.3 that the required exponential integrability is not formally proved. Because these three items are exactly the load-bearing inputs that carry the logarithmic exponent and the L^p integrability, the significance of the result depends on supplying those proofs.

major comments (3)
  1. [Appendix B, Remark B.6 and Proposition B.5] Proposition B.5 is the discrete parabolic extension of the Calderón–Zygmund criterion [AKM19, Lemma 7.2], and Remark B.6 states that the adaptation is 'mostly notational (and omitted here)'. This statement is load-bearing: Proposition B.5 is applied in Step 2 of Proposition B.7 to the coarse-grained function f with the scale-dependent error K in (B.7), and it produces the annealed L^p bounds (4.4)–(4.5) which are then converted, via Proposition 4.3, into the Green's matrix estimates (2.21)–(2.22) used throughout Sections 3.5.1–3.5.3. The parabolic discrete setting requires the approximants f_{(s,y),r} to be defined on parabolic cylinders Q_r(s,y), to satisfy the L∞ bound on the same scale, and to control the time variable and the boundary of Q_{2r}(s,y) on all scales; this is not a purely notational matter. An incorrect constant or exponent in this lemma would directly change the value κ = 4d + 10 in Theorem 2. The proof of Proposition B.5, or a precise reference to a statement covering the discrete parabolic case with these hypotheses, must be supplied.
  2. [Section 4.1.2, Proposition 4.11 (with Section 4.3)] Proposition 4.11 is stated with 'proof omitted here', attributed to a combination of Propositions 4.5 and 4.9. This proposition is the only input for the estimate on d^*_1 H_1 in Section 4.3 and hence for the mixed-codifferential bound (3.26) that drives the d = 3 logarithmic power. The mixed-codifferential estimate carries the exponent 2d + 4 in the numerator and the power 2d + 1 in the denominator; a small change in either exponent would alter the final κ, and the product structure in (4.9) involves a single time integral of two heat kernels whose joint behavior under the Langevin dynamics must be controlled, not merely the product of the two individual L^p bounds. The authors should include the derivation of the second estimate in Proposition 4.11, in particular the treatment of the codifferential in the first variable and the dependence on ∥f∥_{L^{2p}}.
  3. [Section 3.2, Remark 3.3; Appendix A] The proof uses throughout — with exponent p ≥ 8, as stated in Remark B.2 and in Section 4.5 — the assertion that exp(±U_x), exp(U_{cos,x}), cosh(U_x) exp(U_{cos,x}) and their products with trigonometric factors are uniformly in L^p(Ω) for p ≥ 8. Remark 3.3 explicitly says that this 'is not formally proved in the article', and Appendix A only establishes that U_x is well defined in L^2 (via the variance bound for U_{2L} − U_L), asserting without proof that the argument 'can be refined' to obtain integrability of exp(8U_x). These uniform L^p bounds enter the Hölder and Cauchy–Schwarz steps in (3.20), (3.28), (3.29) and (4.13); without them, the constants in those estimates depend on x in an uncontrolled way. A proof, or a precise statement of the relevant generalization of [Fun05, Theorem 4.9] to this setting, should be included.
minor comments (4)
  1. [Section 4.3] The parenthetical after (4.10) concedes that the L^2 bound (3.28) 'should be established for the L^4-norm instead' and that this 'can be achieved by increasing the value of β if necessary'; please state the L^4 version explicitly, since the Hölder step in Section 3.5.3 pairs this bound with an L^4 factor.
  2. [Section 5] The lower bound inherits the transversal asymptotics of [DW24, Theorem 1] together with the Dunlop–Newman inequality via the metric-graph limit (Theorems 3–5). Because [DW24] is the authors' own prior work, it would be helpful to spell out exactly which constants and uniformity statements from that paper are used before passing to the infinite-volume limit.
  3. [Throughout (typesetting)] In the version provided for review, several long displays (e.g., around (2.1), (3.19), (3.21)–(3.24), and (B.8)) contain corrupted glyphs that make the algebra impossible to verify by eye; the authors should ensure the final TeX and PDF rendering is clean, since these identities are central to the argument.
  4. [Section 3.4] The Taylor expansions for a − a′ are written with an informal O(1/|x|^{2d−4}) remainder; stating the remainder explicitly via the second-order Taylor formula with the L∞ and L^2 bounds from (3.11) would make the log-exponent bookkeeping easier to audit.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the upper bound is derived, not fitted, and the lower bound reduces via Dunlop-Newman to the authors' prior transversal asymptotics.

full rationale

The proof of Theorem 2 is a genuine derivation whose inputs do not encode the conclusion. For the upper bound, the chain runs: trigonometric identity (1.5); the Fröhlich-Spencer duality (Proposition 3.1) rewriting the Villain cosine correlations as observables (3.5) of the interface model; removal of lower-order terms (Section 3.3); reduction of the target to the two covariance estimates (3.12) and (3.13); and the Helffer-Sjöstrand representation (Proposition 2.19). The decay exponents are inherited from classical single-scale facts (lattice Green's function and gradient decay, Proposition 2.10) and from the annealed heat-kernel regularity of Proposition 4.5, proved in Appendix B; the Green's-matrix bounds (2.20)-(2.22) and the second-order Green's-matrix bounds of Proposition 4.11 then determine the bookkeeping. The logarithmic exponent kappa = 4d+10 is tracked explicitly, and the paper says it does not optimize the exponent; no free parameter is fitted to the |x|^{-2(d-2)} spin-wave prediction, which enters only as motivation. The lower bound concerns a different observable: it follows from the transversal asymptotics (1.2) of the authors' prior paper [DW24] via the Dunlop-Newman inequality (Theorem 5, built on [DN75] and the metric-graph limit [NW19, AHPS21]). Because the input (transversal correlation) and output (parallel correlation) are distinct quantities, and [DW24, Theorem 1] is parameter-free with assumptions that do not include the target bound, this self-citation is independent support rather than circularity. Two passages carry verification risk that is not circularity: Proposition B.5 is a discrete parabolic Calderón-Zygmund criterion whose proof is cited to [AKM19, Lemma 7.2] with the discrete-parabolic adaptation described as 'mostly notational (and omitted here)' (Remark B.6), and Proposition 4.11 states the second-order Green's-matrix estimates with 'proof omitted here.' If either input failed, the upper bound would not follow; these are completeness and correctness concerns, not circular reductions, so they do not raise the circularity score.

Assumptions & free parameters 0 free parameters · 9 assumptions · 0 invented entities

Every listed input is upstream of Theorem 2 and is not paid for in the paper. There are no fitted free parameters: beta is taken large (a stated regime restriction), kappa=4d+10 is derived, and C1, C2 are explicit consequences of the estimates. The upper bound rests on the duality of Proposition 3.1 (imported from [FS82], [Bau] and [DW24]), on the Helffer-Sjöstrand representation (imported from [HS94], [NS97], [GOS01]), and on the annealed regularity of the heat kernel (developed in Appendix B but with one criterion deferred to [AKM19] and one proposition omitted). The lower bound rests on the same authors' prior transversal asymptotics [DW24, Theorem 1] and on the Dunlop-Newman inequality carried over to the Villain model through a metric graph limit ([DN75], [NW19], [AHPS21]).

assumptions (9)
  • domain assumption Low temperature regime: beta >= beta0(d,p) large enough for the cluster-expansion duality (Proposition 3.1) and for small ellipticity contrast of the spatial operator L_spat (Hessian bounds (2.12)-(2.13)).
    The reduction of the Villain model to the interface measure mu_beta and all regularity statements require beta large; the paper does not quantify beta0, it only asserts existence.
  • domain assumption Prior theorem of the same authors [DW24, Theorem 1]: the transversal two-point function and the spin-spin function satisfy (1.2)-(1.3) with decay |x|^{-(d-2)} and explicit constants c1, c2.
    The lower bound in Theorem 2 and Corollary 1.1 (c1 = c2) are derived from these quoted asymptotics; this paper does not re-prove them.
  • domain assumption Dunlop-Newman correlation inequality for the Villain model (Theorem 5), obtained by passing from the XY model via the metric graph limit ([DN75], [NW19], [AHPS21]).
    Section 5: the inequality combined with the transversal lower bound yields c/|x|^{2(d-2)}; the inequality and the metric graph convergence are imported from the cited literature.
  • standard math Helffer-Sjöstrand representation and Lax-Milgram solvability (Propositions 2.17, 2.19), from [HS94], [NS97], [GOS01].
    The identity cov[f,g] = sum_x <d_x f, H(x)> is the engine of Sections 3 and 4; its application to the exponential observables additionally needs the unproved integrability of Remark 3.3.
  • ad hoc to paper Exponential integrability: exp(+-U_x) exp(U_cos,x), cosh(U_x) exp(U_cos,x) lie in L^p for p >= 8 uniformly in x.
    Remark 3.3 says this 'is not formally proved in the article' and defers to [Fun05, Theorem 4.9] and [DW24, Proposition 3.4]; it is load-bearing for the Helffer-Sjöstrand representation and for (3.20), (3.29).
  • domain assumption Annealed L^p regularity of the heat kernel codifferentials, Proposition 4.5 (p=8): ||d*1 d*2 P||_{L^p} <= C (ln t)^{d+2} t^{-(d/2+1)} exp(-|x-y|/(C sqrt(t))), and the companion bounds (4.3)-(4.5).
    Proved in Appendix B only modulo Proposition B.5, whose discrete parabolic version is cited to [AKM19, Lemma 7.2] with the adaptation omitted; it is the source of the Green's matrix decay (2.20)-(2.22).
  • domain assumption Second-order Helffer-Sjöstrand Green's matrix estimates, Proposition 4.11: ||G_sec,f||_{L^p} <= C ||f||_{L^p} / (|x-x1|_+^{2d-2} + |y-y1|_+^{2d-2}) and the mixed codifferential bound.
    The proof is 'omitted here' per Section 4.1; the H1 estimate (3.26) in Section 4.3 uses it directly together with Lemma 2.6 and the Appendix C summation bounds.
  • standard math Standard discrete analysis inputs: lattice Green's function decay (Proposition 2.10, [LL10]), discrete Poincaré lemma (Lemma 2.4, [FS82]), charge summation (Lemma 2.6), convolution bounds (Appendix C).
    Background facts invoked throughout Sections 2-4; proofs are given or cited to standard references.
  • standard math Existence and symmetries of the infinite-volume states: Ginibre inequality for the Villain measure and the thermodynamic limit, translation invariance and symmetry of mu_beta (Proposition 2.13).
    Used to define the correlation functions and to justify (3.14); Proposition 2.13's proof is only sketched ('can be extended without much difficulties').

how reviews work

0 comments
Cite this review

Pith. "Pith review of Parallel spin wave for the Villain model." pith.science (2026). https://pith.science/paper/6LZCYY6Q

@misc{pith2026250704098,
  author       = {Pith},
  title        = {Pith review of: Parallel spin wave for the Villain model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6LZCYY6Q}},
  note         = {Machine review of arXiv:2507.04098}
}
abstract

In this paper, we study the Villain model in $\mathbb{Z}^d$ in dimension $d\geq 3$. It is conjectured, that the parallel correlation function in the infinite volume Gibbs state, i.e., the map $$ x \mapsto \langle \cos\theta(0) \cos\theta(x) \rangle_{\mu_{\mathrm{Vil}, \beta}} -\left( \langle \cos\theta(0) \rangle_{\mu_{\mathrm{Vil}, \beta}} \right)^2, $$ decays like $|x|^{-2(d-2)}$ as $|x| \to \infty$ at low temperature. The results of Bricmont, Fontaine, Lebowitz, Lieb, and Spencer (1981) show that for the related XY model, this correlation decays at least as fast as $|x|^{2-d}$. We prove the optimal upper and lower bounds for the Villain model in $d=3$, up to a logarithmic correction, and also improve the upper bound in general dimensions. Our proof builds upon the approach developed in our previous article, which in turn is inspired by a key observation of Fr\"{o}hlich and Spencer (1982): in the low temperature regime, a combination of duality transformation and renormalisation allows certain properties of the Villain model to be analysed in terms of a (vector-valued) $\nabla \varphi$ interface model. This latter model can be investigated using the Helffer-Sj\"{o}strand representation formula combined with tools of elliptic and parabolic regularity.

Figures

Figures reproduced from arXiv: 2507.04098 by the authors.

Figure 1
Figure 1. The metric graph G6, where G is a 3 × 3 cube in Z 2 The following result, obtained in [NW19] and [AHPS21], shows that the correlation function on a rescaled sequence of the XY model on metric graphs converges to that of the Villain model. Theorem 4. Let G be a finite graph, β > 0 and n ∈ N, let βn = nβ and Gn be the metric graph obtained by G. Then we have limn→∞ ⟨cos θx cos θy⟩XY,βn,Gn = ⟨cos θx cos θy⟩Vil,β,G. Lik… view at source ↗
Figure 2
Figure 2. The four regions used in the proof of the inequality (C.3). to obtain that ∑ z∈Zd ∶ ∣x∣<∣y−z∣≤ ∣y∣ 2 1 ∣x∣ 2d+1 + + ∣y − z∣ 2d+1 + 1 ∣z∣ d−1 + ≤ ∑ z∈Zd ∶ ∣x∣<∣y−z∣≤ ∣y∣ 2 1 ∣y − z∣ 2d+1 + 1 ∣y∣ d−1 + ≤ C ∣y∣ d−1 + ∑ z∈Zd ∶ ∣x∣<∣z∣≤ ∣y∣ 2 1 ∣z∣ 2d+1 + ≤ C ∣y∣ d−1 + ∑ z∈Zd ∶ ∣x∣<∣z∣ 1 ∣z∣ 2d+1 + ≤ C ∣y∣ d−1 + 1 ∣x∣ d+1 + . Step 3: the third region. In this case, we note that there exists c > 0 such that ∣x∣ 2d+1 + + ∣… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 60 canonical work pages

  1. [1]

    Quantitative hydrodynamic limits of the L angevin dynamics for gradient interface models

    Scott Armstrong and Paul Dario. Quantitative hydrodynamic limits of the L angevin dynamics for gradient interface models. Electron. J. Probab. , 29:Paper No. 9, 93, 2024

    work page 2024

  2. [2]

    Depinning in integer-restricted G aussian F ields and BKT phases of two-component spin models

    Michael Aizenman, Matan Harel, Ron Peled, and Jacob Shapiro. Depinning in integer-restricted G aussian F ields and BKT phases of two-component spin models. arXiv preprint arXiv:2110.09498 , 2021

    arXiv 2021

  3. [3]

    Elliptic homogenization from qualitative to quantitative

    Scott Armstrong and Tuomo Kuusi. Elliptic homogenization from qualitative to quantitative. arXiv preprint arXiv:2210.06488 , 2022

    arXiv 2022

  4. [4]

    Armstrong, T

    S. Armstrong, T. Kuusi, and J.-C. Mourrat. Mesoscopic higher regularity and subadditivity in elliptic homogenization. Comm. Math. Phys. , 347(2):315--361, 2016

    work page 2016

  5. [5]

    The additive structure of elliptic homogenization

    Scott Armstrong, Tuomo Kuusi, and Jean-Christophe Mourrat. The additive structure of elliptic homogenization. Invent. Math. , 208(3):999--1154, 2017

    work page 2017

  6. [6]

    Quantitative stochastic homogenization and large-scale regularity , volume 352

    Scott Armstrong, Tuomo Kuusi, and Jean-Christophe Mourrat. Quantitative stochastic homogenization and large-scale regularity , volume 352. Springer, 2019

    work page 2019

  7. [7]

    D. G. Aronson. Bounds for the fundamental solution of a parabolic equation. Bull. Amer. Math. Soc. , 73:890--896, 1967

    work page 1967

  8. [8]

    Armstrong and Charles K

    Scott N. Armstrong and Charles K. Smart. Quantitative stochastic homogenization of convex integral functionals. Ann. Sci. \'Ec. Norm. Sup\'er. (4) , 49(2):423--481, 2016

    work page 2016

Show all 69 references
  1. [9]

    Local limit theorems for the random conductance model and applications to the Ginzburg--Landau interface model

    Sebastian Andres and Peter A Taylor. Local limit theorems for the random conductance model and applications to the Ginzburg--Landau interface model. Journal of Statistical Physics , 182(2):35, 2021

    2021

  2. [10]

    C^2 regularity of the surface tension for the interface model

    Scott Armstrong and Wei Wu. C^2 regularity of the surface tension for the interface model. Comm. Pure Appl. Math. , 75(2):349--421, 2022

    2022

  3. [11]

    The scaling limit of the continuous solid-on-solid model

    Scott Armstrong and Wei Wu. The scaling limit of the continuous solid-on-solid model. arXiv preprint arXiv:2310.13630 , 2023

    2023 arXiv

  4. [12]

    Bauerschmidt

    R. Bauerschmidt. Ferromagnetic spin systems. lecture notes available at https://cims.nyu.edu/ bauerschmidt/teaching/math253x/spin.pdf

  5. [13]

    Percolation transition for random forests in d 3

    Roland Bauerschmidt, Nicholas Crawford, and Tyler Helmuth. Percolation transition for random forests in d 3 . Invent. Math. , 237(2):445--540, 2024

    2024

  6. [14]

    Bricmont, J

    J. Bricmont, J. R. Fontaine, and L. J. Landau. On the uniqueness of the equilibrium state for plane rotators. Comm. Math. Phys. , 56(3):281--296, 1977

    1977

  7. [15]

    Lebowitz, Elliott H

    Jean Bricmont, Jean-Raymond Fontaine, Joel L. Lebowitz, Elliott H. Lieb, and Thomas Spencer. Lattice systems with a continuous symmetry. III . L ow temperature asymptotic expansion for the plane rotator model. Comm. Math. Phys. , 78(4):545--566, 1980/81

    1980

  8. [16]

    Phase coexistence of gradient G ibbs states

    Marek Biskup and Roman Koteck \'y . Phase coexistence of gradient G ibbs states. Probability theory and related fields , 139(1):1--39, 2007

    2007

  9. [17]

    Herm Jan Brascamp and Elliott H. Lieb. On extensions of the B runn- M inkowski and P r\' e kopa- L eindler theorems, including inequalities for log concave functions, and with an application to the diffusion equation. J. Functional Analysis , 22(4):366--389, 1976

    1976

  10. [18]

    Low temperature properties for correlation functions in classical N -vector spin models

    Tadeusz Balaban and Michael O'Carroll. Low temperature properties for correlation functions in classical N -vector spin models. Comm. Math. Phys. , 199(3):493--520, 1999

    1999

  11. [19]

    The discrete G aussian model, I

    Roland Bauerschmidt, Jiwoon Park, and Pierre-Fran cois Rodriguez. The discrete G aussian model, I . R enormalisation group flow at high temperature. Ann. Probab. , 52(4):1253--1359, 2024

    2024

  12. [20]

    The discrete G aussian model, II

    Roland Bauerschmidt, Jiwoon Park, and Pierre-Fran cois Rodriguez. The discrete G aussian model, II . I nfinite-volume scaling limit at high temperature. Ann. Probab. , 52(4):1360--1398, 2024

    2024

  13. [21]

    Scaling limit for a class of gradient fields with nonconvex potentials

    Marek Biskup and Herbert Spohn. Scaling limit for a class of gradient fields with nonconvex potentials. Ann. Probab. , 39(1):224--251, 2011

    2011

  14. [22]

    Decay of covariances, uniqueness of ergodic component and scaling limit for a class of systems with non-convex potential

    Codina Cotar and Jean-Dominique Deuschel. Decay of covariances, uniqueness of ergodic component and scaling limit for a class of systems with non-convex potential. Ann. Inst. Henri Poincar\' e Probab. Stat. , 48(3):819--853, 2012

    2012

  15. [23]

    Strict convexity of the free energy for a class of non-convex gradient models

    Codina Cotar, Jean-Dominique Deuschel, and Stefan M \"u ller. Strict convexity of the free energy for a class of non-convex gradient models. Communications in mathematical physics , 286(1):359--376, 2009

    2009

  16. [24]

    Validity of the spin-wave approximation for the free energy of the H eisenberg ferromagnet

    Michele Correggi, Alessandro Giuliani, and Robert Seiringer. Validity of the spin-wave approximation for the free energy of the H eisenberg ferromagnet. Comm. Math. Phys. , 339(1):279--307, 2015

    2015

  17. [25]

    Wilson loops in I sing lattice gauge theory

    Sourav Chatterjee. Wilson loops in I sing lattice gauge theory. Comm. Math. Phys. , 377(1):307--340, 2020

    2020

  18. [26]

    A strong central limit theorem for a class of random surfaces

    Joseph G Conlon and Thomas Spencer. A strong central limit theorem for a class of random surfaces. Communications in Mathematical Physics , 325:1--15, 2014

    2014

  19. [27]

    Hydrodynamic limit for a class of degenerate convex -interface models

    Paul Dario. Hydrodynamic limit for a class of degenerate convex -interface models. arXiv preprint arXiv:2409.20405 , 2024

    2024 arXiv

  20. [28]

    Delmotte and J.-D

    T. Delmotte and J.-D. Deuschel. On estimating the derivatives of symmetric diffusions in stationary random environment, with applications to interface model. Probab. Theory Related Fields , 133(3):358--390, 2005

    2005

  21. [29]

    Entropic repulsion for massless fields

    Jean-Dominique Deuschel and Giambattista Giacomin. Entropic repulsion for massless fields. Stochastic processes and their applications , 89(2):333--354, 2000

    2000

  22. [30]

    Large deviations and concentration properties for interface models

    Jean-Dominique Deuschel, Giambattista Giacomin, and Dmitry Ioffe. Large deviations and concentration properties for interface models. Probability theory and related fields , 117:49--111, 2000

    2000

  23. [31]

    Dyson, Elliott H

    Freeman J. Dyson, Elliott H. Lieb, and Barry Simon. Phase transitions in quantum spin systems with isotropic and nonisotropic interactions. J. Statist. Phys. , 18(4):335--383, 1978

    1978

  24. [32]

    Fran ç ois Dunlop and Charles M. Newman. Multicomponent field theories and classical rotators. Comm. Math. Phys. , 44(3):223--235, 1975

    1975

  25. [33]

    A Ray--Knight theorem for interface models and scaling limits

    Jean-Dominique Deuschel and Pierre-Fran c ois Rodriguez. A Ray--Knight theorem for interface models and scaling limits. Probability Theory and Related Fields , 189(1):447--499, 2024

    2024

  26. [34]

    Massless phases for the V illain model in d 3

    Paul Dario and Wei Wu. Massless phases for the V illain model in d 3 . Ast\'erisque , (447):viii+217, 2024

    2024

  27. [35]

    Freeman J. Dyson. General theory of spin-wave interactions. Phys. Rev. (2) , 102:1217--1230, 1956

    1956

  28. [36]

    Kosterlitz- T houless transition line for the two dimensional C oulomb gas

    Pierluigi Falco. Kosterlitz- T houless transition line for the two dimensional C oulomb gas. Comm. Math. Phys. , 312(2):559--609, 2012

    2012

  29. [37]

    Critical exponents of the two dimensional C oulomb gas at the B erezinskii- K osterlitz- T houless transition

    Pierluigi Falco. Critical exponents of the two dimensional C oulomb gas at the B erezinskii- K osterlitz- T houless transition. arXiv preprint arXiv:1311.2237 , 2013

    2013 arXiv

  30. [38]

    u rg Fr\

    J\" u rg Fr\" o hlich and Thomas Spencer. Kosterlitz- T houless transition in the two-dimensional plane rotator and C oulomb gas. Phys. Rev. Lett. , 46(15):1006--1009, 1981

    1981

  31. [39]

    u rg Fr\

    J\" u rg Fr\" o hlich and Thomas Spencer. Massless phases and symmetry restoration in abelian gauge theories and spin systems. Comm. Math. Phys. , 83(3):411--454, 1982

    1982

  32. [40]

    Funaki and H

    T. Funaki and H. Spohn. Motion by mean curvature from the G inzburg- L andau interface model. Comm. Math. Phys. , 185(1):1--36, 1997

    1997

  33. [41]

    Fr\" o hlich, B

    J. Fr\" o hlich, B. Simon, and Thomas Spencer. Infrared bounds, phase transitions and continuous symmetry breaking. Comm. Math. Phys. , 50(1):79--95, 1976

    1976

  34. [42]

    Stochastic interface models

    Tadahisa Funaki. Stochastic interface models. In Lectures on P robability T heory and S tatistics , volume 1869 of Lecture Notes in Math. , pages 103--274. Springer, Berlin, 2005

    2005

  35. [43]

    J. Ginibre. General formulation of G riffiths' inequalities. Comm. Math. Phys. , 16:310--328, 1970

    1970

  36. [44]

    Gloria, S

    A. Gloria, S. Neukamm, and F. Otto. Quantification of ergodicity in stochastic homogenization: optimal bounds via spectral gap on G lauber dynamics. Invent. Math. , 199(2):455--515, 2015

    2015

  37. [45]

    Gloria and F

    A. Gloria and F. Otto. An optimal variance estimate in stochastic homogenization of discrete elliptic equations. Ann. Probab. , 39(3):779--856, 2011

    2011

  38. [46]

    Gloria and F

    A. Gloria and F. Otto. An optimal error estimate in stochastic homogenization of discrete elliptic equations. Ann. Appl. Probab. , 22(1):1--28, 2012

    2012

  39. [47]

    The corrector in stochastic homogenization: optimal rates, stochastic integrability, and fluctuations

    Antoine Gloria and Felix Otto. The corrector in stochastic homogenization: optimal rates, stochastic integrability, and fluctuations. arXiv preprint arXiv:1510.08290 , 2015

    2015 arXiv

  40. [48]

    Quantitative results on the corrector equation in stochastic homogenization

    Antoine Gloria and Felix Otto . Quantitative results on the corrector equation in stochastic homogenization. J. Eur. Math. Soc. (JEMS) , 19(11):3489--3548, 2017

    2017

  41. [49]

    Low-temperature asymptotic expansion for classical O ( N ) vector models

    Alessandro Giuliani and S\'ebastien Ott. Low-temperature asymptotic expansion for classical O ( N ) vector models. Probab. Math. Phys. , 6(2):439--477, 2025

    2025

  42. [50]

    Goldstone

    J. Goldstone. Field theories with ``superconductor'' solutions. Nuovo Cimento (10) , 19:154--164, 1961

    1961

  43. [51]

    Equilibrium fluctuations for interface model

    Giambattista Giacomin, Stefano Olla, and Herbert Spohn. Equilibrium fluctuations for interface model. Ann. Probab. , 29(3):1138--1172, 2001

    2001

  44. [52]

    Improved spin-wave estimate for W ilson loops in U(1) lattice gauge theory

    Christophe Garban and Avelio Sep\'ulveda. Improved spin-wave estimate for W ilson loops in U(1) lattice gauge theory. Int. Math. Res. Not. IMRN , (21):18142--18198, 2023

    2023

  45. [53]

    Trudinger

    David Gilbarg and Neil S. Trudinger. Elliptic partial differential equations of second order . Classics in Mathematics. Springer-Verlag, Berlin, 2001. Reprint of the 1998 edition

    2001

  46. [54]

    Alan H. Guth. Existence proof of a nonconfining phase in four-dimensional U (1) \ lattice gauge theory. Phys. Rev. D (3) , 21(8):2291--2307, 1980

    1980

  47. [55]

    On the correlation for K ac-like models in the convex case

    Bernard Helffer and Johannes Sj\" o strand. On the correlation for K ac-like models in the convex case. J. Statist. Phys. , 74(1-2):349--409, 1994

    1994

  48. [56]

    Spontaneous symmetry breakdown in the abelian H iggs model

    Tom Kennedy and Chris King. Spontaneous symmetry breakdown in the abelian H iggs model. Comm. Math. Phys. , 104(2):327--347, 1986

    1986

  49. [57]

    Ordering, metastability and phase transitions in two-dimensional systems

    John Michael Kosterlitz and David James Thouless. Ordering, metastability and phase transitions in two-dimensional systems. Journal of Physics C: Solid State Physics , 6(7):1181, 1973

    1973

  50. [58]

    Random walk: a modern introduction , volume 123

    Gregory F Lawler and Vlada Limic. Random walk: a modern introduction , volume 123. Cambridge University Press, 2010

    2010

  51. [59]

    Fluctuations for the G inzburg- L andau interface model on a bounded domain

    Jason Miller. Fluctuations for the G inzburg- L andau interface model on a bounded domain. Comm. Math. Phys. , 308(3):591--639, 2011

    2011

  52. [60]

    Absence of ferromagnetism or antiferromagnetism in one-or two-dimensional isotropic H eisenberg models

    N David Mermin and Herbert Wagner. Absence of ferromagnetism or antiferromagnetism in one-or two-dimensional isotropic H eisenberg models. Physical Review Letters , 17(22):1133, 1966

    1966

  53. [61]

    Naddaf and T

    A. Naddaf and T. Spencer. On homogenization and scaling limit of some gradient perturbations of a massless free field. Comm. Math. Phys. , 183(1):55--84, 1997

    1997

  54. [62]

    Newman and Wei Wu

    Charles M. Newman and Wei Wu. Lee- Y ang property and G aussian multiplicative chaos. Comm. Math. Phys. , 369(1):153--170, 2019

    2019

  55. [63]

    Sheffield

    S. Sheffield. Random surfaces. Ast\' e risque , (304):vi+175, 2005

    2005

  56. [64]

    Sj \"o strand

    J. Sj \"o strand. Correlation asymptotics and W itten L aplacians. Algebra i Analiz , 8(1):160--191, 1996

    1996

  57. [65]

    Euclidean proof of the G oldstone T heorem

    Kurt Symanzik. Euclidean proof of the G oldstone T heorem. Communications in Mathematical Physics , 6(3):228--232, 1967

    1967

  58. [66]

    Symanzik

    K. Symanzik. Renormalizable models with simple symmetry breaking. I . S ymmetry breaking by a source term. Comm. Math. Phys. , 16:48--80, 1970

    1970

  59. [67]

    Localization and delocalization of random interfaces

    Yvan Velenik. Localization and delocalization of random interfaces. Probab. Surv. , 3:112--169, 2006

    2006

  60. [68]

    Theory of one-and two-dimensional magnets with an easy magnetization plane

    Jacques Villain. Theory of one-and two-dimensional magnets with an easy magnetization plane. II . T he planar, classical, two-dimensional magnet. Journal de Physique , 36(6):581--590, 1975

    1975

  61. [69]

    Local central limit theorem for gradient field models

    Wei Wu. Local central limit theorem for gradient field models. Probab. Theory Related Fields , 191(1-2):1--40, 2025

    2025

Pith tools

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