Convergence of form factor expansions in 1+1-dimensional integrable quantum field theories beyond space-like two-point functions of sinh-Gordon

The form-factor (bootstrap) construction represents correlation functions of a 1+1-dimensional integrable quantum field theory as series over the number of intermediate particles, and its rigorous convergence is known only for space-like separated two-point functions of the sinh-Gordon model, under a growth condition on the operator-dependent functions $p_N^{(O)}$ (Kozlowski, Inventiones Mathematicae 2023: $|\mathcal{U}_N(r)|\le\exp[-(3\pi^4b\hat b/4)N^2/(\ln N)^3(1+O(1/\ln N))]$, whence uniform convergence). The source itself formulates the completion of this program as open: (i) time-like separations, requiring a deformation of the rapidity contours to curves $\gamma(u)=u+i\vartheta(u)$; (ii) multipoint correlation functions; (iii) the sine-Gordon model, whose form factors involve multiple particle species and bound states; (iv) the full two-scale asymptotic expansion of the N-particle summand $\mathcal{U}_N(r)$ in $N$ and $\ln N$ beyond the upper bound; and (v) other operators of physical interest beyond those explicitly treated. Every listed direction remains open: multipoint convergence is conjectural, the Wightman axiomatics of sinh-Gordon is proven only conditionally on it, and the sine-Gordon form factors were written down explicitly only in 2026, with their convergence proofs deferred to future work. The problem is to carry out these extensions, turning the forty-year-old bootstrap construction into a rigorous constructive method.

Background

An integrable quantum field theory in 1+1 dimensions is a massive relativistic model with infinitely many higher-spin conserved charges; this symmetry is strong enough to force scattering to be elastic and factorized, so that every N-particle process is a product of two-body phases $S(\theta_i-\theta_j)$, where $\theta$ is the rapidity parametrizing one-dimensional relativistic momentum ($p^0=m\cosh\theta$, $p^1=m\sinh\theta$). The sinh-Gordon model — the theory of one real scalar field with an exponential $\cosh(b\varphi)$ self-interaction, and the simplest affine Toda field theory — is the simplest case: its spectrum is a single self-conjugated massive particle with no bound states, and its two-particle S-matrix is a unitary pure phase depending on one coupling $b\in(0,1/2)$.

The form factor bootstrap is the standard non-perturbative method for such theories. A local operator $O$ is characterized by its form factors, the matrix elements $F_N^{O}(\theta_1,\dots,\theta_N)$ between the vacuum and N-particle in-states, which are determined by solving a closed system of functional equations (Watson equations expressing the effect of the S-matrix on analytic continuation in rapidity, crossing symmetry, and residue conditions at kinematic poles). Correlation functions are then represented as form factor series: a two-point Wightman function is a sum over the number $N$ of intermediate particles, $\sum_N\mathcal{U}_N(r)$, where $\mathcal{U}_N(r)$ is an N-fold rapidity integral built from the squared moduli of the form factors and the two-particle phase. Physicists have used this construction for four decades, but its convergence was essentially uncontrolled: before the source paper, the only rigorous convergence proof for any model was an unpublished note treating a single two-point function in the Lee-Yang model (the integrable quantum field theory with the simplest non-trivial factorized S-matrix).

The source paper (Kozlowski, arXiv:2007.01740; Inventiones Mathematicae 2023) established the first general result. Under a growth condition on the operator-dependent functions $p_N^{(O)}$ appearing in the form factors, schematically $|p_N^{(O)}(\beta|\ell)|\le C_1^N\prod_a e^{C_2|\beta_a|^s}$ for constants $C_1,C_2,s$ — a hypothesis satisfied by the natural operator families (exponentials of the field, conserved currents, components of the energy-momentum tensor) — it proves that the N-particle summand obeys $|\mathcal{U}_N(r)|\le\exp[-(3\pi^4b\hat b/4)\,N^2/(\ln N)^3\,(1+O(1/\ln N))]$ with $\hat b=1/2-b$, whence the series converges uniformly for space-like separations $r\ge r_0>0$. Structurally the proof analyses $\mathcal{U}_N(r)$ much as one analyses a $\beta$-ensemble partition function in random-matrix theory — an N-fold integral of the exponential of a pairwise-interaction energy between eigenvalue-like variables (the paper draws this analogy explicitly; $\mathcal{U}_N(r)$ is the N-th summand of the two-point series).

Space-like separation of the two insertion points makes the integrand amenable to direct estimates; at time-like separation an oscillatory phase appears, and the contour of each rapidity integration must be deformed to a curve $\gamma(u)=u+i\vartheta(u)$ in the strip $|\mathrm{Im}\,\theta|<\pi$, a step the source describes as involving additional technical — but not conceptual — complications that it does not carry out. The Wightman axioms are the standard axioms (relativistic covariance, locality, spectral condition, existence and uniqueness of the vacuum) that correlation functions must satisfy for a genuine quantum field theory to exist; a 2025 companion paper proves they hold for the sinh-Gordon bootstrap construction conditionally on convergence of the form-factor series, which makes the present problem the blocking ingredient for an unconditional construction of the model.

Two further ingredients make the open directions concrete. First, the sine-Gordon model, the integrable theory of one real scalar with interaction $\cos\varphi$, has a richer spectrum: a soliton/antisoliton doublet carrying topological charge and, for suitable coupling, infinitely many neutral bound-state breathers, so its form factors carry species indices and must satisfy additional bound-state residue conditions. Second, the source conjectures that the true large-$N$ behaviour of $\mathcal{U}_N(r)$ is a two-scale asymptotic expansion of the shape $\ln\mathcal{U}_N(r)\simeq-N^2(\ln N)^p\{Q_{0,0}+\sum_{(s,\ell)\neq(0,0)}Q_{s,\ell}/(N^\ell(\ln N)^s)\}$ with the power $p$ and the coefficients $Q_{s,\ell}$ left undetermined; only the upper bound above is proven, and the author describes the appearance of the $1/(\ln N)^3$ prefactor as subtle, with no heuristic explanation.

Problem Statement

Complete the convergence program for the form-factor (bootstrap) expansion initiated by the source: establish rigorous convergence of the series — and the sharp large-$N$ asymptotics of its summands — in the settings left open, namely: (1) Time-like regime: prove convergence of the sinh-Gordon two-point form-factor series for time-like separated points, carrying out the required deformation of the rapidity contours to curves $\gamma(u)=u+i\vartheta(u)$ and controlling the analytic structure crossed by the deformation. (2) Multipoint functions: prove that the k-point correlation functions of the sinh-Gordon model admit convergent form-factor series representations — in particular establish the conjecture of arXiv:2502.03894 that the $\mathbf{r}$-truncated k-point representations — obtained by bounding, for each $j=1,\dots,k-1$, the number of particles in the intermediate channel between the $j$-th and $(j+1)$-st field insertion by $r_j$, the full k-point function arising upon summing over $\mathbf{r}=(r_1,\dots,r_{k-1})$ — converge to the full k-point Wightman functions. (3) Sine-Gordon: extend the convergence theory to the sine-Gordon model, whose form factors involve multiple particle species (soliton, antisoliton, breathers) and bound-state poles. (4) Sharp asymptotics: replace the upper bound on $\mathcal{U}_N(r)$ by a full two-scale asymptotic expansion in $N$ and $\ln N$, determining the power $p$ and the coefficients $Q_{s,\ell}$, or disprove the conjectured shape. (5) Operator class: extend the convergence result to at least one further class of local operators of physical interest beyond those explicitly treated in the source — namely the exponentials of the field, the conserved currents, and the components of the energy-momentum tensor — exhibiting that class and establishing the convergence of its form-factor series; resolving this direction requires settling at least one such further class and does not require a full classification of all admissible operators. These five directions are the open extensions posed by the source and its follow-ups, all serving one program: turning the bootstrap construction of 1+1-dimensional integrable quantum field theories from factorized scattering data into a rigorous method. Directions (1)-(2), combined with the conditional results already available, would yield an unconditional Wightman construction of sinh-Gordon; (3) extends it to the first model with bound states; (4) supplies the sharp quantitative control the method currently lacks; (5) widens its operator scope. Completion standard: a full resolution of the task must establish all five directions jointly, and each direction is an independent open problem whose rigorous settlement is a partial resolution of the program rather than a full solution; conjectural or heuristic arguments do not by themselves resolve a direction.

References

1. Karol K. Kozlowski, On convergence of form factor expansions in the infinite volume quantum Sinh-Gordon model in 1+1 dimensions, Inventiones Mathematicae 233 (2023) 725-827, DOI 10.1007/s00222-023-01192-7, arXiv:2007.01740, https://arxiv.org/abs/2007.01740
2. Karol Kajetan Kozlowski, Bootstrap approach to 1+1-dimensional integrable quantum field theories: the case of the sinh-Gordon model, Proceedings of the International Congress of Mathematicians 2022, DOI 10.4171/ICM2022/96, arXiv:2112.14274, https://arxiv.org/abs/2112.14274
3. Karol K. Kozlowski and Alex Simon, On multipoint correlation functions in the Sinh-Gordon 1+1 dimensional quantum field theory, arXiv:2502.03894 (2025), https://arxiv.org/abs/2502.03894
4. Karol K. Kozlowski and Alex Simon, Wightman axiomatics of the bootstrap construction of the 1+1 dimensional Sinh-Gordon model, arXiv:2511.10800 (2025), https://arxiv.org/abs/2511.10800
5. Karol K. Kozlowski and Alex Simon, Form factors of field exponentials in the 1+1 dimensional Sine-Gordon model, arXiv:2608.14365 (2026), https://arxiv.org/abs/2608.14365
6. Nikolay Barashkov, Joona Oikarinen, and Mo Dick Wong, Small deviations of Gaussian multiplicative chaos and the free energy of the two-dimensional massless Sinh-Gordon model, arXiv:2408.16574 (2024), https://arxiv.org/abs/2408.16574
7. Colin Guillarmou, Trishen S. Gunaratnam, and Vincent Vargas, 2d Sinh-Gordon Model on the Infinite Cylinder, Communications in Mathematical Physics 407 (2026) 97, DOI 10.1007/s00220-026-05588-3, arXiv:2405.04076, https://arxiv.org/abs/2405.04076
8. Wojciech Dybalski, Exact Schwinger Functions for a Class of Bounded Interactions in d >= 2, Communications in Mathematical Physics 406 (2025) 211, DOI 10.1007/s00220-025-05391-6, arXiv:2502.07546, https://arxiv.org/abs/2502.07546
9. Charlie Dworaczek Guera and Karol K. Kozlowski, On the equilibrium measure for the Lukyanov integral, Journal of Mathematical Physics 66 (2025) 082301, DOI 10.1063/5.0252653, arXiv:2406.00145, https://arxiv.org/abs/2406.00145
10. Alice Guionnet, Karol Kozlowski, and Alex Little, Asymptotic expansion of the partition function for beta-ensembles with complex potentials, arXiv:2411.10610 (2024), https://arxiv.org/abs/2411.10610
11. Constantin Babenko, Frank Göhmann, Karol K. Kozlowski, and Junji Suzuki, A thermal form factor series for the longitudinal two-point function of the Heisenberg-Ising chain in the antiferromagnetic massive regime, Journal of Mathematical Physics 62 (2021) 041901, DOI 10.1063/5.0039863, arXiv:2011.12752, https://arxiv.org/abs/2011.12752
12. Frank Göhmann, Karol K. Kozlowski, Jesko Sirker, and Junji Suzuki, Spin conductivity of the XXZ chain in the antiferromagnetic massive regime, SciPost Physics 12 (2022) 158, DOI 10.21468/SciPostPhys.12.5.158, arXiv:2202.05304, https://arxiv.org/abs/2202.05304