In the repulsive regime of the sine-Gordon model, finite-volume vacuum expectation values of local fields can be expressed, through the fermionic (hidden-Grassmann) basis, in terms of finitely many elements of an infinite matrix whose UV (small-volume) series is built from the kink nonlinear integral equation (NLIE) describing the finite-volume spectrum. Their ultraviolet limit encodes diagonal three-point couplings of the model's UV conformal field theory, a perturbed complex Liouville theory. Hegedus computed the first two coefficients of this UV series (the leading one for all operator parameters, the second only at vanishing operator parameter), and Hegedus and Roth subsequently conjectured a closed-form determinant expression for the leading UV coefficient valid for all integer operator indices, verified numerically to about 19 significant digits. Both results rest on an unproven assumption: UV-expansion terms whose volume-power scalings are not predicted by the UV field theory are assumed to cancel or vanish, an assumption supported only by high-precision numerics. The open problem is to turn this conjectural picture into an analytic derivation: prove the required cancellations, establish the all-index leading formula as a theorem, and extend analytic control to the subleading (second-order) coefficient at general operator parameter, with a treatment uniform in the coupling, including the delicate regime 1<p<31<p<3.

Background

The sine-Gordon model is a two-dimensional relativistic quantum field theory of a scalar boson with a cosine self-interaction of coupling β\beta; it is integrable (exactly solvable), with factorized soliton scattering and a finite-volume spectrum governed by the Destri-de Vega nonlinear integral equation (NLIE), a nonlinear integral equation for a counting function that encodes the Bethe quantization conditions on a cylinder of circumference LL. In the repulsive regime, parameterized by p=β2/(1−β2)>1p=\beta^{2}/(1-\beta^{2})>1 (equivalently ν=1/(p+1)∈(0,1/2)\nu=1/(p+1)\in(0,1/2)), the model is best regarded as a perturbation of complex Liouville conformal field theory (Liouville theory at imaginary coupling), whose central charge is cL=1−6ν2/(1−ν)c_{L}=1-6\nu^{2}/(1-\nu). As L→0L\to 0 (equivalently the dimensionless volume ℓ=ML\ell=\mathcal{M}L, with M\mathcal{M} the soliton mass, tends to zero), the theory flows back to this ultraviolet (UV) conformal field theory (CFT).

By the state-operator correspondence of two-dimensional CFT, the matrix elements of local fields on the cylinder scale with the three-point couplings (OPE coefficients, also called structure constants) of the UV CFT: a diagonal matrix element ⟨O∣Φ∣O⟩\langle\mathcal{O}|\Phi|\mathcal{O}\rangle determines a three-point coupling in which two of the three operators coincide. A program initiated by Bajnok and Janik extracts such CFT data from the integrable description of the massive theory, and for the sine-Gordon model the key tool is the fermionic basis discovered by Jimbo, Miwa and Smirnov: expectation values of local fields are compactly represented in a Grassmann-odd operator basis, and every finite-volume expectation value reduces to finitely many elements ω2k−1,1−2j(α)\omega_{2k-1,1-2j}(\alpha) of an infinite matrix, where α∈[0,2)\alpha\in[0,2) labels the vertex operator being averaged and an additional twist parameter αz\alpha_{z} in the NLIE encodes the sandwiching state. Hegedus (2020) showed how these matrix elements are computed from the NLIE spectral data for arbitrary sandwiching states.

The UV limit of the ω\omega-matrix is obtained by expanding around an asymptotic solution of the NLIE constructed from the kink functions Z±(x)Z_{\pm}(x), i.e. the NLIE solutions on the two soliton-kink backgrounds, which approach common plateau values at large argument with exponentially small corrections (plateau expansions). The expansion runs in powers of ℓ4/(p+1)\ell^{4/(p+1)}, and delicate constant-part (CP) prescriptions extract finite constants from ℓ\ell-dependent integrals, with analytic subtleties that depend on the regime of pp (notably for 1<p<31<p<3). Perturbed-CFT arguments predict that physical expectation values may contain only specific powers of ℓ\ell; however, the term-by-term UV expansion of the ω\omega-matrix also generates contributions with other, unpredicted ℓ\ell-power scalings. The existing computations assume these unwanted contributions vanish or cancel among themselves — an assumption supported only by high-precision numerics — and this assumption is precisely what blocks the existing results from being theorems. The current state is: the leading-order UV coefficient of the ratio ⟨Φα+2(1−ν)/ν⟩/⟨Φα⟩\langle\Phi_{\alpha+2(1-\nu)/\nu}\rangle/\langle\Phi_{\alpha}\rangle of primary-field vacuum expectation values is known for all 1<p<∞1<p<\infty, 0≤α<20\le\alpha<2; the second-order coefficient is known only at α=0\alpha=0; and a conjectural closed-form determinant formula for the leading UV coefficient for every integer operator index mm in ⟨Φα+2m(1−ν)/ν⟩/⟨Φα⟩\langle\Phi_{\alpha+2m(1-\nu)/\nu}\rangle/\langle\Phi_{\alpha}\rangle (Hegedus-Roth) is verified numerically to about 19 digits but not derived.

Problem Statement

Within the integrable formulation of finite-volume expectation values of the sine-Gordon model in its repulsive regime (p>1p>1, ν=1/(p+1)\nu=1/(p+1), dimensionless volume ℓ=ML\ell=\mathcal{M}L), vacuum expectation values of the primary fields Φα\Phi_{\alpha} are expressed through finitely many elements of the infinite matrix ω2k−1,1−2j(α)\omega_{2k-1,1-2j}(\alpha), whose ℓ→0\ell\to 0 series is generated by expanding around the asymptotic (kink) solution of the NLIE. Field theory predicts that physical expectation values contain only specific powers of ℓ\ell, with exponents that depend on the operator parameters and on the matrix indices: the ratios ⟨Φα+2m(1−ν)/ν⟩/⟨Φα⟩\langle\Phi_{\alpha+2m(1-\nu)/\nu}\rangle/\langle\Phi_{\alpha}\rangle scale as ℓ−2m(m−(m+1)ν)−2mαν(1+O(ℓ4/(p+1)))\ell^{-2m(m-(m+1)\nu)-2m\alpha\nu}(1+O(\ell^{4/(p+1)})), and the matrix elements ω2k−1,1−2j(α)\omega_{2k-1,1-2j}(\alpha) as ℓ−2(j+k−1)+2(2−α)/(p+1)\ell^{-2(j+k-1)+2(2-\alpha)/(p+1)} times a series in integer powers of ℓ4/(p+1)\ell^{4/(p+1)} (Hegedűs JHEP 01 (2025) 084, eqs. (4.20), (4.22), (4.23); Hegedűs-Roth arXiv:2606.20018, eqs. (2.24), (4.5)), yet the term-by-term expansion of the ω\omega-matrix also produces contributions with other ℓ\ell-power scalings, which all existing computations assume cancel or vanish — an assumption supported only by high-precision numerics. The problem is to replace this assumption-based picture by a complete analytic derivation, with the following jointly defining content:

  1. Prove, within the NLIE framework (kink functions, their plateau expansions, the linear integral equations for the coefficient functions, the constant-part prescriptions and the ±i0\pm i0 contour conventions, as pinned in Hegedűs JHEP 01 (2025) 084 and Hegedűs-Roth arXiv:2606.20018), that all contributions to the UV expansion of the ω\omega-matrix whose ℓ\ell-power scalings are not of the field-theory-predicted form cancel or vanish in the physical vacuum expectation values, uniformly for p>1p>1 (including the regime 1<p<31<p<3 where the constant-part prescriptions differ) and for 0≤α<20\le\alpha<2, and that only the 'effective' sector of contributions enters the leading coefficient.

  2. Consequently derive, rather than conjecture, the leading UV coefficient C(α,m,ν)\mathcal{C}(\alpha,m,\nu) of the ratios of vacuum expectation values ⟨Φα+2m(1−ν)/ν⟩/⟨Φα⟩\langle\Phi_{\alpha+2m(1-\nu)/\nu}\rangle/\langle\Phi_{\alpha}\rangle for every integer m≥1m\ge 1 and 0≤α<20\le\alpha<2 — in particular establish (or correct) the conjectured closed-form determinant formula (5.12) of Hegedűs-Roth expressing C(α,m,ν)\mathcal{C}(\alpha,m,\nu) through the plateau-expansion coefficients of the kink functions and the integrals of the kink linear equations.

  3. Extend the analytic control to the subleading order: derive the explicit second-order coefficient of the UV series of the ω\omega-matrix at general α∈[0,2)\alpha\in[0,2) (presently computed only at α=0\alpha=0), with a uniform account of the constant-part prescriptions across the parameter range.

The problem is completed when the three items above are established throughout the stated parameter range: the required cancellations proven rather than assumed, the leading UV coefficient C(α,m,ν)\mathcal{C}(\alpha,m,\nu) derived for every integer m≥1m\ge 1 and 0≤α<20\le\alpha<2, and the second-order coefficient derived at general α∈[0,2)\alpha\in[0,2). Extensions to other models (N=1 supersymmetric sine-Gordon, sinh-Gordon), to the special integer-pp twist values corresponding to the minimal models Vir(p,p+1), and to non-vacuum sandwiching states are declared future directions in the sources and are not part of this problem's completion boundary.

References

  1. Árpád Hegedűs, "Integrable formulation of the OPE coefficients in the UV limit of the sine-Gordon model", JHEP 01 (2025) 084, doi:10.1007/JHEP01(2025)084, arXiv:2410.16883, https://arxiv.org/abs/2410.16883
  2. Árpád Hegedűs and Apor Roth, "Leading UV Formula for Finite-Volume Vertex Operator Expectation Values in the Sine-Gordon Model from Kink NLIE", arXiv:2606.20018, https://arxiv.org/abs/2606.20018
  3. Zoltan Bajnok and Romuald A. Janik, "OPE coefficients and the mass-gap from the integrable scattering description of 2D CFT's", JHEP 11 (2022) 128, doi:10.1007/JHEP11(2022)128, arXiv:2209.10393, https://arxiv.org/abs/2209.10393
  4. M. Jimbo, T. Miwa and F. Smirnov, "Hidden Grassmann structure in the XXZ model V: sine-Gordon model", Lett. Math. Phys. 96 (2011) 325-365, doi:10.1007/s11005-010-0438-9, arXiv:1007.0556, https://arxiv.org/abs/1007.0556
  5. Árpád Hegedűs, "Finite volume expectation values in the sine-Gordon model", JHEP 01 (2020) 122, doi:10.1007/JHEP01(2020)122, arXiv:1909.08467, https://arxiv.org/abs/1909.08467
  6. G. Feverati, F. Ravanini and G. Takács, "Non-linear integral equation and finite volume spectrum of minimal models perturbed by \Phi_{(1,3)}", Nucl. Phys. B 570 (2000) 615-643, doi:10.1016/S0550-3213(99)00771-3, arXiv:hep-th/9909031, https://arxiv.org/abs/hep-th/9909031