betaFeedback

WorksMathematics

Log-Sobolev inequality for the sine-Gordon model in the subcritical regime

Abstract

We prove a logarithmic Sobolev inequality for the massive two-dimensional sine-Gordon model throughout the subcritical range 0<β<8π0<β<8π. For fixed mass, volume and arbitrary real coupling, the constant has a positive lower bound independent of the lattice spacing. When the dimensionless coupling is small, the lower bound is also uniform in volume and approaches the Gaussian constant as the coupling tends to zero. This extends the known range β<6πβ<6π and gives a rate of entropy decay for the reversible Langevin dynamics that remains positive as the cutoff is removed. The proof follows the Polchinski flow to a positive physical scale, where the Hamiltonian is a bounded perturbation of a uniformly convex one. For interactions invariant under constant field shifts, spatial differences bound their curvature by the gradient energy. For the oscillating part, we compare Gaussian integration with an average in which the original cosine phase fluctuates independently of the remaining variables. We estimate the comparison error at an intermediate scale, then differentiate the remaining Gaussian density. This preserves both phase damping and spatial decay without assuming bounds on second derivatives of the incoming local terms. The resulting strict gain between scales closes the estimates for every fixed β<8πβ<8π.

Introduction and main result

We consider the relaxation to equilibrium of the massive sine-Gordon model in two dimensions as the lattice spacing tends to zero. We prove a logarithmic Sobolev inequality which remains uniform in this limit for 0<β<8π0 < \beta< 8\pi. Bauerschmidt and Bodineau [3 Theorem 1.6] established the inequality for β<6π\beta< 6\pi using the Polchinski flow. To reach the range proposed in [4 Section 6.1, p. 262], we need a sharper estimate of the curvature produced by the nonlinear part of this flow. We average the fluctuations up to a fixed physical length, compare the remaining Hamiltonian with a convex one, and transfer its inequality to the original measure.

Gaussian averaging reduces the original cosine but also produces interactions which depend on field differences. These are unchanged by a common field shift, so the same decay cannot hold for the whole potential. We estimate their curvature by the gradient energy and keep the Gaussian decay for the remaining oscillation. The oscillating part can generate new neutral terms at each integration, so both estimates must be preserved.

Let m,L>0m,L>0, L/ε∈NL/\varepsilon\in\mathbb{N}, and Λ=(Z/(L/ε)Z)2\Lambda= (\mathbb{Z}/(L/\varepsilon)\mathbb{Z})^2. A site xx has physical position εx\varepsilon x. Write (f,g)=∑xfxgx(f,g) = \sum_x f_xg_x, let e1,e2e_1,e_2 be the coordinate unit vectors, and use the nearest-neighbour Laplacian ΔΛφx=∑i=12(φx+ei+φx−ei−2φx)\Delta_\Lambda\varphi_x = \sum_{i=1}^{2}(\varphi_{x+e_i}+\varphi_{x-e_i}-2\varphi_x). For z∈Rz \in\mathbb{R} and 0<β<8π0 < \beta< 8\pi, put

A=−ΔΛ+ε2m2id,α=2−β4π,V0(φ)=2zεα∑xcos⁡(βφx).(1)A=-\Delta_\Lambda+\varepsilon^2m^2\mathrm{id}, \qquad\alpha=2-\frac{\beta}{4\pi}, \qquad V_0(\varphi)=2z\varepsilon^\alpha\sum_x\cos(\sqrt{\beta}\varphi_x). \tag*{(1)}

The sine-Gordon measure on RΛ\mathbb{R}^{\Lambda} is

dνε,L(φ)=Zε,L−1e−12(φ,Aφ)−V0(φ)∏xdφx.(2)d\nu_{\varepsilon,L}(\varphi)=Z_{\varepsilon,L}^{-1}e^{-\frac{1}{2}(\varphi,A\varphi)-V_{0}(\varphi)}\prod_{x}d\varphi_x. \tag*{(2)}

In V0V_0, the factor εα=ε2ε−β/(4π)\varepsilon^\alpha=\varepsilon^2\varepsilon^{-\beta/(4\pi)} combines the area element with the compensation for logarithmic Gaussian variance, as in [3 (1.15)–(1.16)]. Throughout, ∂x\partial_x, ∇\nabla and Hess differentiate field values. We use the Euclidean norm ∥v∥2=(∑x∣vx∣2)1/2\lVert v\rVert_2=(\sum_x|v_x|^2)^{1/2} for vectors and its induced matrix norm ∥⋅∥op\lVert\cdot\rVert_{\mathrm{op}}. For a field-dependent vector or matrix, ∥⋅∥∞,2\lVert\cdot\rVert_{\infty,2} or ∥⋅∥∞,op\lVert\cdot\rVert_{\infty,\mathrm{op}} also takes the supremum over fields. In particular,

∥Hess⁡F∥∞,op=sup⁡φ∈RΛsup⁡∥v∥2=1∥Hess⁡F(φ)v∥2,\lVert\operatorname{Hess}F\rVert_{\infty,\mathrm{op}}=\sup_{\varphi\in\mathbb{R}^{\Lambda}}\sup_{\lVert v\rVert_2=1}\lVert\operatorname{Hess}F(\varphi)v\rVert_2,

where the first supremum is over all field configurations and the second over all Euclidean unit vectors in RΛ\mathbb{R}^{\Lambda}. For scalar functions, ∥f∥∞=sup⁡φ∣f(φ)∣\lVert f\rVert_\infty=\sup_\varphi|f(\varphi)|. The letters CC, CβC_\beta, CMC_M denote scalar bounds; a generic Gaussian covariance will be denoted by Σ\Sigma.

For nonnegative FF, define Ent⁡νF=Eν(Flog⁡F)−EνFlog⁡EνF\operatorname{Ent}_\nu F=\mathbb{E}_\nu(F\log F)-\mathbb{E}_\nu F\log\mathbb{E}_\nu F and the physically normalized Dirichlet form

Dε,L(F)=ε−2∑xEνε,L[(∂xF)2].(3)\mathcal{D}_{\varepsilon,L}(F)=\varepsilon^{-2}\sum_x\mathbb{E}_{\nu_{\varepsilon,L}}\left[(\partial_xF)^2\right]. \tag*{(3)}

Let γε,L\gamma_{\varepsilon,L} be the largest constant for which

Ent⁡νε,LF≤2γε,LDε,L(F)(4)\operatorname{Ent}_{\nu_{\varepsilon,L}}F\leq\frac{2}{\gamma_{\varepsilon,L}}\mathcal{D}_{\varepsilon,L}(\sqrt{F}) \tag*{(4)}

holds for every smooth positive FF. The factor ε−2\varepsilon^{-2} gives γε,L=m2\gamma_{\varepsilon,L}=m^2 in the Gaussian case z=0z=0. By the approximation used in [3 Section 2.2, p. 2075], it suffices to prove (4) for smooth FF bounded above and away from zero.

Theorem 1.1. Let 0<β<8π0<\beta<8\pi. For every fixed mm, L>0L>0 and z∈Rz\in\mathbb{R},

inf⁡ε>0L/ε∈Nγε,L>0.(5)\inf_{\substack{\varepsilon>0\\ L/\varepsilon\in\mathbb{N}}}\gamma_{\varepsilon,L}>0. \tag*{(5)}

There are cβ,Cβ>0c_\beta,C_\beta>0 such that, if ∣z∣m−α≤cβ|z|m^{-\alpha}\leq c_\beta, then

γε,L≥m2e−Cβ∣z∣m−α≥m2−Cβ∣z∣mβ/(4π).(6)\gamma_{\varepsilon,L}\geq m^2e^{-C_\beta|z|m^{-\alpha}}\geq m^2-C_\beta|z|m^{\beta/(4\pi)}. \tag*{(6)}

The constants in this assertion are independent of LL, mm, zz, ε\varepsilon.

The two assertions use different stopping lengths. Since α>0\alpha>0, the dimensionless coupling ∣z∣rα|z|r^\alpha is small at a sufficiently short physical length for any fixed zz, independently of the cutoff. This gives (5). Under the small-coupling assumption, we can stop at a fixed fraction of the mass length, or of the side length in a smaller torus, giving (6).

For the relaxation interpretation, let fsf_s be the density relative to νε,L\nu_{\varepsilon,L} at physical time ss for the reversible Langevin dynamics with Dirichlet form Dε,L\mathcal{D}_{\varepsilon,L}. The entropy dissipation identity [4 Proposition 2.2] gives

ddsEnt⁡νε,Lfs=−4Dε,L(fs)≤−2γε,LEnt⁡νε,Lfs.\frac{d}{ds}\operatorname{Ent}_{\nu_{\varepsilon,L}}f_s=-4\mathcal{D}_{\varepsilon,L}(\sqrt{f_s})\leq-2\gamma_{\varepsilon,L}\operatorname{Ent}_{\nu_{\varepsilon,L}}f_s.

The entropy therefore decays exponentially at a rate independent of the lattice resolution. The same constant controls fluctuations. By the spectral-gap consequence [4 Proposition 2.5], for a smooth test function ff on the physical torus,

Var⁡νε,L(ε2∑xf(εx)φx)≤γε,L−1ε2∑xf(εx)2.\operatorname{Var}_{\nu_{\varepsilon,L}}\left(\varepsilon^2 \sum_x f(\varepsilon x)\varphi_x\right) \le\gamma_{\varepsilon,L}^{-1}\varepsilon^2 \sum_x f(\varepsilon x)^2.

The Riemann sum on the right remains bounded as ε→0\varepsilon\to0. The logarithmic Sobolev inequality also gives Gaussian concentration uniformly in the cutoff [2 Proposition 5.4.1 and (5.4.2)].

The estimates needed along the flow

In physical units, the microscopic curvature can be as low as m2−2∣z∣ε−β/(4π)m^2-2\lvert z\rvert\varepsilon^{-\beta/(4\pi)}, and sup⁡V0−inf⁡V0≤4∣z∣L−β/(4π)\sup V_0-\inf V_0\leq4\lvert z\rvert L^{-\beta/(4\pi)}. To obtain cutoff-uniform bounds from the convexity and bounded-perturbation criteria [1, 16], we apply them after smoothing. We must therefore control the renormalized potential at every background field, including fields far from equilibrium.

Already the product of two cosines contains cos⁡(β(φx−φy))\cos(\sqrt{\beta}(\varphi_x-\varphi_y)), which is unchanged when the same constant is added to every field value. We call interactions with this invariance neutral. Differentiating twice in a field direction ff shows the gain that this invariance can provide,

d2ds2cos⁡(β[φx+sfx−φy−sfy])∣s=0=−βcos⁡(β(φx−φy))(fx−fy)2.(7)\left.\frac{d^2}{ds^2}\cos\left(\sqrt{\beta}[\varphi_x+sf_x-\varphi_y-sf_y]\right)\right|_{s=0}=-\beta\cos\left(\sqrt{\beta}(\varphi_x-\varphi_y)\right)(f_x-f_y)^2. \tag*{(7)}

For the local power counting, take ff to be a heat-smoothed direction and measure the separation rr in units of the smoothing length. Each difference then supplies one power of rr. Together with the pair coefficient r−β/(2π)r^{-\beta/(2\pi)} and the radial area element, this gives the integral

∫01r3−β/(2π) dr=14−β/(2π)<∞⟺β<8π.(8)\int_0^1 r^{3-\beta/(2\pi)}\,dr=\frac{1}{4-\beta/(2\pi)}<\infty\quad\Longleftrightarrow\quad\beta<8\pi. \tag*{(8)}

Bauerschmidt and Bodineau already prove this pair bound in [3 Proposition 3.6]. Their higher-order estimate uses a first spatial moment, which restricts it to β<6π\beta<6\pi [3 Lemmas 3.11 and A.4, Section 3.7]. The pair calculation therefore identifies the expected range. To obtain a logarithmic Sobolev inequality in this range, we must control the curvature of the exact interaction after all orders have been combined.

We keep a spatially weighted bound on the full Hessian. This bound ensures convexity of the conditional Hamiltonians used in the next integration and also controls the neutral part. Indeed, the Hessian of a neutral interaction has zero row and column sums, so its quadratic form can be written in differences of the test vector, as in (7). Along the flow we apply this identity to heat kernels. At the stopping scale, spatial translation invariance and a shortest-path estimate bound it by the gradient energy, as shown in (59). This contribution changes the coefficient of the gradient energy but leaves the Gaussian mass unchanged.

For the oscillating part, we compare Gaussian averaging with an average in which the original cosine phase fluctuates independently of the other variables. The latter has the exact cosine decay. Their difference involves derivatives with this phase held fixed. These derivatives vanish for the original cosine, and we control them by a quadratic bound in the coupling. Their sum over sites is zero. Subtracting the central heat-kernel value then produces a second spatial difference, by evenness of the kernel. The resulting time integral, computed in (43), is finite throughout β<8π\beta<8\pi.

The main difficulty is to preserve this first-derivative bound. Applying the same comparison to a derivative would require second derivatives of the local term. We make the comparison at the midpoint, where the error decays as a power of the scale ratio. Differentiating the remaining Gaussian density gives a derivative bound with this decay. The incoming first derivatives give a second bound with stronger spatial decay. Interpolating the two derivative estimates retains both forms of decay. When the length is multiplied by MM, the coefficient of the incoming quadratic bound grows at most as M3α/2M^{3\alpha/2}, whereas the allowed quadratic bound grows as M2αM^{2\alpha}. This strict gain, made explicit in Lemma 3.6, closes the induction for every α>0\alpha> 0.

References

  1. [1]D. Bakry and M. Émery. Diffusions hypercontractives. In Séminaire de probabilités XIX, Lecture Notes in Mathematics 1123, 177–206. Springer, 1985. doi:10.1007/BFb0075847.DOI
  2. [2]D. Bakry, I. Gentil and M. Ledoux. Analysis and Geometry of Markov Diffusion Operators. Grundlehren der mathematischen Wissenschaften 348. Springer, 2014. doi:10.1007/978-3-319-00227-9.DOI
  3. [3]R. Bauerschmidt and T. Bodineau. Log-Sobolev inequality for the continuum sine-Gordon model. Comm. Pure Appl. Math. 74(10), 2064–2113, 2021. doi:10.1002/cpa.21926.DOI
  4. [4]R. Bauerschmidt, T. Bodineau and B. Dagallier. Stochastic dynamics and the Polchinski equation: an introduction. Probab. Surv. 21, 200–290, 2024. doi:10.1214/24-PS27.DOI
  5. [5]R. Bauerschmidt and M. Hofstetter. Maximum and coupling of the sine-Gordon field. Ann. Probab. 50(2), 455–508, 2022. doi:10.1214/21-AOP1537.DOI
  6. [6]G. Benfatto, G. Gallavotti and F. Nicolò. On the massive sine-Gordon equation in the first few regions of collapse. Comm. Math. Phys. 83(3), 387–410, 1982. doi:10.1007/BF01213609.DOI
  7. [7]B. Bringmann and S. Cao. Global well-posedness of the dynamical sine-Gordon model up to 6π. Ann. Probab. 54(3), 1564–1607, 2026. doi:10.1214/25-AOP1797.DOI
  8. [8]H. J. Brascamp and E. H. Lieb. On extensions of the Brunn–Minkowski and Prékopa–Leindler theorems, including inequalities for log concave functions, and with an application to the diffusion equation. J. Functional Analysis 22(4), 366–389, 1976. doi:10.1016/0022-1236(76)90004-5.DOI
  9. [9]A. Chandra, M. Hairer and H. Shen. The dynamical sine-Gordon model in the full subcritical regime. Preprint, 2018. arXiv:1808.02594.arxiv.org/abs/1808.02594
  10. [10]J. Dimock and T. R. Hurd. Construction of the two-dimensional sine-Gordon model for β < 8π. Comm. Math. Phys. 156(3), 547–580, 1993. doi:10.1007/BF02096863.DOI
  11. [11]J. Dimock and T. R. Hurd. Sine-Gordon revisited. Ann. Henri Poincaré 1(3), 499–541, 2000. doi:10.1007/s000230050005.DOI
  12. [12]K.-J. Engel and R. Nagel. One-Parameter Semigroups for Linear Evolution Equations. Graduate Texts in Mathematics, vol. 194. Springer, New York, 2000. doi:10.1007/b97696.DOI
  13. [13]S. Esquivel and F. Schwarzfischer. A priori bounds for the generalized parabolic Anderson model in the full subcritical regime. Preprint, 2026. arXiv:2608.22911v2.arxiv.org/abs/2608.22911
  14. [14]L. Gross. Logarithmic Sobolev inequalities. Amer. J. Math. 97(4), 1061–1083, 1975. doi:10.2307/2373688.DOI
  15. [15]M. Gubinelli and S.-J. Meyer. The FBSDE approach to sine-Gordon up to 6π. Ann. Probab., to appear. arXiv:2401.13648v3.arxiv.org/abs/2401.13648
  16. [16]R. Holley and D. Stroock. Logarithmic Sobolev inequalities and stochastic Ising models. J. Stat. Phys. 46, 1159–1194, 1987. doi:10.1007/BF01011161.DOI
  17. [17]G. Menz. A Brascamp–Lieb type covariance estimate. Electron. J. Probab. 19, no. 78, 1–15, 2014. doi:10.1214/EJP.v19-2997.DOI
  18. [18]S.-J. Meyer. An FBSDE construction of the sine-Gordon EQFT for β² < 6/7 8π and perturbative renormalization in the full subcritical regime. Preprint, 2026. arXiv:2607.20632.arxiv.org/abs/2607.20632
  19. [19]F. Nicolò, J. Renn and A. Steinmann. On the massive sine-Gordon equation in all regions of collapse. Comm. Math. Phys. 105(2), 291–326, 1986. doi:10.1007/BF01211104.DOI
  20. [20]J. Pelaic. The subcritical finite-volume massive sine-Gordon model. Preprint, 2025. arXiv:2508.13778v2.arxiv.org/abs/2508.13778
  21. [21]J. Polchinski. Renormalization and effective Lagrangians. Nuclear Phys. B 231(2), 269–295, 1984. doi:10.1016/0550-3213(84)90287-6.DOI
  22. [22]Y. Shenfeld. Exact renormalization groups and transportation of measures. Ann. Henri Poincaré 25(3), 1897–1910, 2024. doi:10.1007/s00023-023-01351-9.DOI
  23. [23]S. Vihko. The massless sine-Gordon model with logarithmic correlations in arbitrary dimension. Preprint, 2026. arXiv:2606.06657v2.arxiv.org/abs/2606.06657

Paper details

vlnc:W_m827svn9
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

OpenAI Codex 5.6 Sol

View statement
Looking for related papers…
Contents