The generalized chiral Potts models (gCPM) are the paradigmatic exactly-solved lattice models whose R-matrix lacks the rapidity-difference property and whose spectral parameter lives on an algebraic curve of genus at least two. Moosavian-Yamazaki-Zhou (2025) engineered the gCPM inside the 4d Chern-Simons theory with a fractional-power meromorphic one-form, explaining both peculiarities and tying the model to hyperbolic SU(n) monopoles, but left the R-matrix itself uncomputed: the one-form has zeros on the spectral curve, so the effective loop-counting parameter diverges, the perturbative expansion that normally produces classical and quantum R-matrices is unavailable, and the gCPM has no classical R-matrix to begin with. The open problem is to compute the gCPM R-matrix from first principles in the 4d Chern-Simons path integral: give a non-perturbative (or demonstrably computable) definition of the theory with the prescribed one-form, construct line defects carrying the Nn−1N^{n-1}-dimensional minimal cyclic representations of Uq(sl^(n,C))U_q(\widehat{\mathfrak{sl}}(n,\mathbb{C})) at qN=1q^N=1, show that the crossing of two such line defects reproduces the known gCPM R-matrix (the intertwiner of minimal cyclic representations, equivalently the star-triangle weights), and determine the relation between ℏ\hbar and the root-of-unity parameters (n,N)(n,N).

Background

Two-dimensional integrable lattice models are statistical (or quantum) spin systems whose row-to-row transfer matrices commute, which is what makes them exactly solvable. Commutativity follows from the Yang-Baxter equation (YBE), an algebraic consistency condition on the Boltzmann weights encoded in an operator R(z,z′)\mathcal{R}(z,z') - the R-matrix - that intertwines two auxiliary spaces labelled by a 'spectral parameter' zz. For most exactly-solved models the R-matrix depends only on the difference of spectral parameters (the 'rapidity-difference property'), and zz ranges over a genus-zero or genus-one curve, giving the familiar rational, trigonometric, and elliptic solution classes.

The chiral Potts model, the ZN\mathbb{Z}_N-symmetric generalization of the Ising model discovered in 1987-88, is the paradigmatic exception: its Boltzmann weights depend on two independent points of a high-genus algebraic curve (genus larger than one already for N>2N>2), and its R-matrix, obtained from the star-triangle relation (a local re-summation identity for Boltzmann weights on the triangular lattice, equivalent to the Yang-Baxter equation) by Baxter, Perk, and Au-Yang (1988), has no rapidity-difference form. Date, Jimbo, Miki, and Miwa (1991) identified the generalized chiral Potts models (gCPM) associated with SU(n)\mathrm{SU}(n) - with n−1n-1 spins each taking values in ZN\mathbb{Z}_N - with the representation theory of the quantum affine algebra Uq(sl^(n,C))U_q(\widehat{\mathfrak{sl}}(n,\mathbb{C})) (a q-deformation of the loop algebra of sl(n)\mathfrak{sl}(n); more precisely, of a trivial extension thereof, as in the Date-Jimbo-Miki-Miwa paper) at a root of unity: the R-matrix is the intertwiner of tensor products of 'minimal cyclic representations', which are Nn−1N^{n-1}-dimensional representations (with N≥3N\geq 3 odd, qN=1q^N = 1, and continuous parameters z∈(C×)3n−1z\in(\mathbb{C}^{\times})^{3n-1}) on which the standard Chevalley generators (the eie_i, fif_i, kik_i of the quantum affine algebra) act by cyclic shifts; the existence of an intertwiner R(z,z′)\mathcal{R}(z,z') forces (z,z′)(z,z') onto the spectral curve of the model.

The 4d Chern-Simons (CS) theory introduced by Costello, and applied to lattice models by Costello and Yamazaki, provides a gauge-theoretic machine for constructing R-matrices: it is a holomorphic Chern-Simons theory on C×P1\mathbb{C}\times\mathbb{P}^1 (a two-dimensional 'topological' plane C\mathbb{C} times the spectral-parameter sphere) whose action is weighted by a meromorphic one-form ω\omega on P1\mathbb{P}^1. Line defects (Wilson lines) running along the two directions and carrying representations of the quantum affine algebra cross at a point, and their expectation value ⟨LhLv⟩\langle\mathbf{L}_h\mathbf{L}_v\rangle computes the R-matrix. When ω\omega has only poles, this is a perturbative expansion in the loop-counting parameter ℏ\hbar whose first non-trivial term is the classical R-matrix r(z)∈End(g⊗g)r(z)\in\mathrm{End}(\mathfrak{g}\otimes\mathfrak{g}); this machinery produces precisely the difference-type (rational/trigonometric/elliptic) R-matrices.

Moosavian, Yamazaki, and Zhou (2025) connected the two stories. Building on an old observation of Atiyah relating hyperbolic monopoles to the Yang-Baxter equation, and on the construction of Murray and Singer of the spectral curves of non-integral hyperbolic monopoles (Bogomolny monopoles, i.e. solutions of B=DΦB=D\Phi for a gauge field and Higgs field on hyperbolic three-space, whose Higgs field approaches non-integral limits at infinity), they extended the correspondence to SU(n)\mathrm{SU}(n) and then engineered the gCPM inside the 4d CS theory: representing the genus-≥2\geq 2 spectral curve as a branched cover of P1\mathbb{P}^1 and requiring topological invariance along C\mathbb{C} together with ZNn−1\mathbb{Z}_N^{n-1}-invariance almost uniquely fixes the fractional-power one-form

ωP1=dz(∏i=12Nn−1(z−zQi))1/Nn−1.\omega_{\mathbb{P}^1}=\frac{\mathsf{d}z}{\left(\prod_{i=1}^{2N^{n-1}}(z-z_{Q_i})\right)^{1/N^{n-1}}}.

This explains the gCPM's peculiarities (higher-genus spectral curve, absence of rapidity difference) as consequences of the branched, fractional-power structure. The construction is, however, at the level of classical engineering: the R-matrix is not computed. On the spectral curve, the pullback of ω\omega vanishes at the ramification points with positive order (the paper derives the order-N−1N-1 vanishing for the chiral Potts case n=2n=2; its gCPM pullback formula (5.42), obtained from the chiral Potts one (5.24) by N→Nn−1N\to N^{n-1}, gives order Nn−1−1N^{n-1}-1, whereas its ramification data (5.40) - 2Nn−12N^{n-1} ramification points each of index (N−1)(n−1)+1(N-1)(n-1)+1 - implies order (N−1)(n−1)(N-1)(n-1), the two agreeing only at n=2n=2 within the paper's explicitly heuristic §5.2 generalization), and a zero of ω\omega makes the effective ℏ\hbar diverge, so the standard perturbative expansion is invalid; consistently, the gCPM has no classical R-matrix that could serve as its leading term. Prior attempts at a non-perturbative formulation of 4d CS - string-theoretic realizations, and the brane/categorification proposal of Ashwinkumar, Tan, and Zhao, whose recipe the authors call formal and not useful for practical computations - do not supply a workable computational scheme.

Problem Statement

Compute the R-matrix of the generalized chiral Potts model directly from the 4d Chern-Simons theory, as posed by Moosavian-Yamazaki-Zhou (sections 5.1 and 7 of arXiv:2502.17545): make the formal path integral

⟨LhLv⟩∼∫DA  LhLv  exp⁡(−12πℏ∫C×P1ωP1∧CS(A))\langle\mathbf{L}_h\mathbf{L}_v\rangle\sim\int\mathcal{D}A\;\mathbf{L}_h\mathbf{L}_v\;\exp\left(-\frac{1}{2\pi\hbar}\int_{\mathbb{C}\times\mathbb{P}^1}\omega_{\mathbb{P}^1}\wedge\mathrm{CS}(A)\right)

with the fractional-power one-form ωP1=dz/(∏i=12Nn−1(z−zQi))1/Nn−1\omega_{\mathbb{P}^1}=\mathsf{d}z/\left(\prod_{i=1}^{2N^{n-1}}(z-z_{Q_i})\right)^{1/N^{n-1}} a well-defined and practically computable object, and evaluate it for two crossing line defects so as to reproduce the known gCPM R-matrix. This comprises three linked components:

  1. (Non-perturbative definition) Provide a non-perturbative definition - or a demonstrably computable equivalent - of the 4d CS path integral with the above one-form. The usual perturbative ℏ\hbar-expansion is unavailable: the one-form has zeros on the spectral curve (vanishing at the ramification points with positive order - the source's pullback formula (5.42) gives Nn−1−1N^{n-1}-1, while its ramification data (5.40) implies (N−1)(n−1)(N-1)(n-1) for n>2n>2), which drives the effective ℏ→∞\hbar\to\infty, and the gCPM has no classical R-matrix to serve as the leading term. Routes mentioned by the source include a string-theoretic realization of the theory or a definition through dimensional reduction of the 6d/10d holomorphic Chern-Simons theories.

  2. (Roots of unity and line defects) Construct, within this 4d CS framework, line defects (Wilson lines or equivalent) carrying the Nn−1N^{n-1}-dimensional minimal cyclic representations of Uq(sl^(n,C))U_q(\widehat{\mathfrak{sl}}(n,\mathbb{C})) at qN=1q^N=1 (N≥3N\geq 3 odd), and control their non-perturbative interactions.

  3. (The R-matrix and the ℏ\hbar-(n,N)(n,N) relation) Show that the crossing expectation value of the two line defects equals the known gCPM R-matrix R(z,z′)\mathcal{R}(z,z') - the intertwiner of the tensor product of two minimal cyclic representations, whose existence confines (z,z′)(z,z') to the gCPM spectral curve - for general n≥2n\geq 2 and odd N≥3N\geq 3, and determine the exact relation between the loop-counting parameter ℏ\hbar of the path integral and the root-of-unity parameters (n,N)(n,N) at qN=1q^N=1 (the limiting regimes N→∞N\to\infty at fixed nn; N,n→∞N,n\to\infty at fixed n/Nn/N; and N,n→∞N,n\to\infty with n/N→0n/N\to 0 are optional aids, not required for completion).

Components 1 and 2 are prerequisites of component 3; a complete answer computes the R-matrix (component 3) from a controlled definition of the theory (component 1) via the explicit line-defect construction (component 2). The same source lists further companion open directions - whether integrable models exist for mixed cyclic spin groups ZN1×⋯×ZNn−1\mathbb{Z}_{N_1}\times\cdots\times\mathbb{Z}_{N_{n-1}} with distinct NiN_i and for the general fractional-power one-forms of its equations (7.2)-(7.6), non-maximal symmetry breaking, gCPM-type models for other semisimple Lie algebras, integral (circle-invariant) hyperbolic monopoles, a string-theory embedding, and the rigorous justification of the branch-cut and distributional manipulations - which also remain open but are not required for completion of the core objective above.

References

  1. Seyed Faroogh Moosavian, Masahito Yamazaki, Yehao Zhou, "Hyperbolic Monopoles, (Semi-)Holomorphic Chern-Simons Theories, and Generalized Chiral Potts Models", arXiv:2502.17545 (2025), DOI 10.48550/arXiv.2502.17545, https://arxiv.org/abs/2502.17545.
  2. Kevin Costello, Masahito Yamazaki, "Gauge Theory And Integrability, III", arXiv:1908.02289 (2019), https://arxiv.org/abs/1908.02289.
  3. R. J. Baxter, J. H. H. Perk, H. Au-Yang, "New solutions of the star-triangle relations for the chiral Potts model", Phys. Lett. A 128 (1988) 138-142, DOI 10.1016/0375-9601(88)90896-1, https://doi.org/10.1016/0375-9601(88)90896-1.
  4. Etsuro Date, Michio Jimbo, Kei Miki, Tetsuji Miwa, "Generalized chiral Potts models and minimal cyclic representations of Uq(gl^(n,C))U_q(\widehat{\mathfrak{gl}}(n,\mathbb{C}))", Commun. Math. Phys. 137 (1991) 133-147, DOI 10.1007/BF02099119, https://doi.org/10.1007/BF02099119.
  5. Michael Atiyah, "Magnetic monopoles and the Yang-Baxter equations", Int. J. Mod. Phys. A 06 (1991) 2761-2774, DOI 10.1142/S0217751X91001349, https://doi.org/10.1142/S0217751X91001349.
  6. Michael Murray, Michael Singer, "Spectral curves of non-integral hyperbolic monopoles", Nonlinearity 9 (1996) 973-997, DOI 10.1088/0951-7715/9/4/009, https://doi.org/10.1088/0951-7715/9/4/009.
  7. Meer Ashwinkumar, Meng-Chwan Tan, Qin Zhao, "Branes and Categorifying Integrable Lattice Models", arXiv:1806.02821 (2018), published as Adv. Theor. Math. Phys. 24 (2020) 1-24, DOI 10.4310/ATMP.2020.v24.n1.a1, https://arxiv.org/abs/1806.02821.
  8. Roland Bittleston, Lionel Mason, Seyed Faroogh Moosavian, "The Self-Duality Equations on a Riemann Surface and Four-Dimensional Chern-Simons Theory", arXiv:2601.05309 (2026), https://arxiv.org/abs/2601.05309.
  9. Eric Vernier, Yuan Miao, Masahito Yamazaki, "Lattice non-invertible symmetry from non-commuting transfer matrices", arXiv:2606.25660 (2026), https://arxiv.org/abs/2606.25660.
  10. Zhao Zhang, "The Yang-Baxter Equation for the Chiral Potts Model and Integrable Parafermions", arXiv:2606.11146 (2026), https://arxiv.org/abs/2606.11146.
  11. Paul Sutcliffe, "Hyperbolic monopole data", arXiv:2507.14990 (2025), published as JHEP 02 (2026) 128, DOI 10.1007/JHEP02(2026)128, https://arxiv.org/abs/2507.14990.
  12. HE YiJun, CHEN Bin, "Four-dimensional Chern-Simons theory and integrable models", Sci. Sin. Phys. Mech. Astron. 56 (2026) 230002, DOI 10.1360/SSPMA-2025-0168, https://doi.org/10.1360/SSPMA-2025-0168.
  13. Sylvain Lacroix, Nat Levine, Anders Wallberg, "1-loop renormalisability of integrable sigma-models from 4d Chern-Simons theory", arXiv:2505.23890 (2025), https://arxiv.org/abs/2505.23890.
  14. Meer Ashwinkumar, Matthias Blau, "Integrable Deformations of the Breitenlohner-Maison Model from 4d Chern-Simons Theory", arXiv:2604.26452 (2026), published as JHEP 07 (2026) 123, DOI 10.1007/JHEP07(2026)123, https://arxiv.org/abs/2604.26452.
  15. Meer Ashwinkumar, Jitendra Pal, "The Yang-Baxter Sigma Model from Twistor Space", arXiv:2602.11288 (2026), published as Phys. Rev. D 114 (2026) 046008, DOI 10.1103/99hb-q8bd, https://arxiv.org/abs/2602.11288.
  16. Meer Ashwinkumar, "R-matrices from Feynman Diagrams in 5d Chern-Simons Theory and Twisted M-theory", arXiv:2408.15732 (2024), https://arxiv.org/abs/2408.15732.