The Komatsu–Mahajan–Shao (KMS) integrability index I(J)=c(J,J)−c(J+1,J−1)\mathcal{I}(J)=c(J,J)-c(J+1,J-1), computed from the operator spectrum at the ultraviolet (UV) fixed point of an asymptotically free two-dimensional coset sigma model, gives a sufficient condition for the survival of spin-JJ conserved currents under quantization and hence for quantum integrability. The 2019 paper that introduced it left one open validity caveat and a set of future directions that frame the open problems for completing this diagnostic program: (1) the invariance of the index, and the Goldschmidt–Witten sufficiency argument it encodes, is established only within conformal perturbation theory around the UV fixed point, and nonperturbative mass-gap effects could generate lower-dimension obstruction (A-type) operators that invalidate the criterion — no example or theorem is known; (2) the index is blind to 't Hooft anomaly data that distinguish integrable points of the same model (e.g., CP1\mathbb{CP}^1 at θ=0\theta=0 versus θ=π\theta=\pi), and the proposed anomaly-refined index based on torus partition functions with symmetry line insertions has not been constructed; (3) the diagnostic concerns only local conserved currents, leaving open how its conclusions relate to integrability carried by non-local charges; (4) the refined index Ir(J)=I(J)+c(2,0)\mathcal{I}_r(J)=\mathcal{I}(J)+c(2,0) proposed for fine-tuned, fermionic, or supersymmetric models remains uncomputed — the sole fermionic extension so far, the plain index of CPN−1\mathbb{CP}^{N-1} coupled to a massless Dirac fermion, found the fermionic contribution vanishing identically; (5) extension to superstring sigma models is open.

Background

In a two-dimensional (1+1-dimensional) quantum field theory, quantum integrability means the presence of an infinite tower of mutually commuting conserved charges. When these charges arise from local conserved currents of Lorentz spin J>2J>2 (operators of dimension equal to their spin, generalizations of the stress tensor built for example from powers of its components), scattering is constrained to be factorized (no particle production), and this underlies the exact solvability of the classic integrable models via the S-matrix bootstrap (exact reconstruction of two-particle scattering amplitudes from consistency conditions) and Bethe-ansatz methods. Many two-dimensional nonlinear sigma models — field theories whose fields parametrize a target space, here a coset space G/HG/H with GG a compact Lie group and HH a subgroup — are classically integrable (their equations of motion can be recast as a zero-curvature, or Lax, representation: the flatness condition of an auxiliary connection), but whether the associated higher-spin currents survive quantization is a delicate, model-dependent question.

Goldschmidt and Witten (1980) gave an operator-counting criterion. A classically conserved spin-JJ current can fail conservation quantum mechanically only through specific operators: 'A-type' operators — conformal primaries (highest-weight operators of the conformal group, as opposed to their derivative descendants) of dimension J+1J+1 and spin J−1J-1 that can appear on the right-hand side of the would-be conservation equation ∂−OJ,J=0\partial_- \mathcal{O}_{J,J}=0 — and 'B-type' operators whose coefficients must be tuned to zero. If the number of candidate currents exceeds the number of obstructions, at least one quantum conserved current is guaranteed to exist.

Komatsu, Mahajan, and Shao (2019) systematized this counting. Coset sigma models are asymptotically free: they become weakly coupled free conformal field theories in the ultraviolet, with global symmetry GG and an operator spectrum computable by representation theory (characters of the global conformal group, combined into a generating partition function by a plethystic exponential (a generating function that packages arbitrary products of the elementary building blocks), with projections implementing discrete symmetries such as charge conjugation). Writing c(Δ,J)c(\Delta,J) for the number of primary operators of dimension Δ\Delta and spin JJ at the UV fixed point, they defined the integrability index I(J)=c(J,J)−c(J+1,J−1)\mathcal{I}(J)=c(J,J)-c(J+1,J-1) and showed it is invariant under conformal perturbation theory away from the fixed point (conserved-current short multiplets can pair with long multiplets, but the two terms shift together). Positivity of I(J)\mathcal{I}(J) for some J>2J>2 is a sufficient condition for quantum integrability; it is not necessary, since finely tuned models can be integrable despite a non-positive index. Applications: the O(N)O(N) model (the sigma model whose target space is the sphere SN−1S^{N-1}) has I(4)=I(6)=1\mathcal{I}(4)=\mathcal{I}(6)=1 once charge conjugation is imposed, establishing a previously unknown spin-6 quantum conserved current; the CPN−1\mathbb{CP}^{N-1} models for N≥3N\ge 3 have all non-positive indices, consistent with their non-integrability; the flag sigma models U(N)/U(1)NU(N)/U(1)^N have all negative indices, making integrability unlikely; and CP1\mathbb{CP}^1 (the O(3)O(3) model) is integrable at θ=0\theta=0 and θ=π\theta=\pi, exactly where an extra Z2\mathbb{Z}_2 charge-conjugation symmetry exists and the projected indices become positive.

The paper was explicit about what its diagnostic does not do, and these gaps define the present problem. First, the invariance argument is a conformal-perturbation-theory statement; asymptotically free sigma models nonperturbatively generate a mass gap in the infrared, and in a massive theory one can in principle write A-type obstruction operators of dimension below J+1J+1, which would invalidate the sufficiency condition — the authors knew of no example. Second, the index is insensitive to 't Hooft anomalies (obstructions to gauging a set of global symmetries simultaneously, detectable in 1+1 dimensions through the anomalous modular transformation of torus partition functions with topological symmetry line insertions, i.e., line operators implementing the symmetries); the CP1\mathbb{CP}^1 model at θ=0\theta=0 (no anomaly; trivially gapped infrared) and at θ=π\theta=\pi (mixed anomaly between SO(3)SO(3) and the Z2\mathbb{Z}_2 charge conjugation; gapless infrared described by the SU(2)1SU(2)_1 Wess–Zumino–Witten theory (a solvable interacting conformal field theory)) are both integrable with identical index values but different anomaly data, and the paper proposed — without constructing — an anomaly-aware refinement. Third, the analysis concerns local currents only, whereas some sigma models owe their integrability to non-local conserved charges (Lüscher-type charges that are not integrals of local densities, protected for example when HH is simple or by a discrete symmetry). Fourth, integrability can be restored by coupling to fermions or by tuning marginal operators (operators of dimension 2): the supersymmetric CPN−1\mathbb{CP}^{N-1} model retains its non-local charge to all orders (Gomes–Abdalla–Abdalla), and the fine-tuned U(1)×SU(N)U(1)\times SU(N) model is integrable (Basso–Rej); the paper introduced a refined index Ir(J)=I(J)+c(2,0)\mathcal{I}_r(J)=\mathcal{I}(J)+c(2,0), adding the number of marginal primaries, as a candidate diagnostic for such fine-tuned integrability, noting it is not sufficient but makes integrability likelier somewhere in parameter space. Fifth, the method is confined to coset sigma models with a UV fixed point of the requisite type, while superstring worldsheet sigma models — for example the Ramond–Ramond (RR) flux backgrounds (RR fluxes: a type of background field strength in string theory) of the AdS/CFT integrable-string program — lie outside it.

Problem Statement

The problem is to resolve the open questions the source paper left for the integrability-index diagnostic of quantum integrability of two-dimensional sigma models: its nonperturbative validity and the future directions listed in its conclusions. The components are independent open directions toward turning the ultraviolet-perturbative sufficient condition I(J)=c(J,J)−c(J+1,J−1)>0\mathcal{I}(J)=c(J,J)-c(J+1,J-1)>0 into a reliable diagnostic:

(1) Nonperturbative validity. The invariance of I(J)\mathcal{I}(J) along the renormalization-group flow, and the Goldschmidt–Witten sufficiency argument it encodes, are established only within conformal perturbation theory around the UV fixed point. Asymptotically free coset sigma models nonperturbatively generate a mass gap, and in a massive theory one can in principle write A-type obstruction operators of dimension below J+1J+1, which would invalidate the sufficiency condition. Determine whether such nonperturbative violations occur: either exhibit a concrete model in which a current counted positively by the index fails to be conserved quantum mechanically through such a nonperturbative obstruction, or prove that they cannot occur (extending the invariance argument beyond conformal perturbation theory).

(2) Anomaly refinement. The index is insensitive to 't Hooft anomaly data that qualitatively distinguish integrable points of the same model — CP1\mathbb{CP}^1 at θ=0\theta=0 (no anomaly, trivially gapped infrared) and θ=π\theta=\pi (mixed SO(3)SO(3)–Z2\mathbb{Z}_2 anomaly, gapless SU(2)1SU(2)_1 infrared) both have I(4)=I(6)=1\mathcal{I}(4)=\mathcal{I}(6)=1. Construct a concrete anomaly-refined index — the source proposes interpreting the index as a torus partition function, possibly with global-symmetry line insertions whose modular transformations encode the anomaly — and validate it where anomaly data are known to affect the integrable flow.

(3) Non-local charges. The diagnostic is built from local conserved currents only. Clarify the relation between a non-positive index conclusion and integrability carried by non-local conserved charges, in particular in marginal or finely-tuned models.

(4) Fermionic, supersymmetric, and fine-tuned models. Compute the refined index Ir(J)=I(J)+c(2,0)\mathcal{I}_r(J)=\mathcal{I}(J)+c(2,0), or an analogous diagnostic, for models whose integrability is restored by coupling to fermions or by tuning marginal couplings — such as the supersymmetric CPN−1\mathbb{CP}^{N-1} model and the fine-tuned U(1)×SU(N)U(1)\times SU(N) models of Basso and Rej, whose integrability is known to arise only at tuned loci of the parameter space — and establish whether the refined index detects such fine-tuned integrability; the source's conjecture is the weaker statement that Ir(J)>0\mathcal{I}_r(J)>0, while not a sufficient condition for quantum integrability, makes integrability more likely to be achieved somewhere in parameter space.

(5) Superstring sigma models. Extend the method to superstring worldsheet sigma models, where new integrable backgrounds could be identified.

Each component is an open question or direction identified by the source paper. They are independent directions rather than a single bundled deliverable; an answer that settles any of them constitutes partial progress toward a reliable diagnostic.

References

  1. Shota Komatsu, Raghu Mahajan, Shu-Heng Shao, "An Index for Quantum Integrability", SciPost Phys. 7, 065 (2019), arXiv:1907.07186, DOI:10.21468/SciPostPhys.7.5.065, https://arxiv.org/abs/1907.07186
  2. Y. Y. Goldschmidt, Edward Witten, "Conservation Laws in Some Two-dimensional Models", Phys. Lett. B 91, 392–396 (1980), DOI:10.1016/0370-2693(80)91004-7, https://inspirehep.net/literature/152203
  3. Jia Tian, Jue Hou, Bin Chen, "Notes on index of quantum integrability", Commun. Theor. Phys. 73, 055001 (2021), arXiv:2002.04952, DOI:10.1088/1572-9494/abe9aa, https://arxiv.org/abs/2002.04952
  4. João Caetano, Shota Komatsu, Yifan Wang, "Large charge 't Hooft limit of N=4\mathcal{N}=4 super-Yang-Mills", JHEP 02, 047 (2024), arXiv:2306.00929, DOI:10.1007/JHEP02(2024)047, https://arxiv.org/abs/2306.00929
  5. Daniele Bielli, Christian Ferko, Liam Smith, Gabriele Tartaglino-Mazzucchelli, "Auxiliary Field Deformations of (Semi-)Symmetric Space Sigma Models", JHEP 01, 096 (2025), arXiv:2409.05704, DOI:10.1007/JHEP01(2025)096, https://arxiv.org/abs/2409.05704
  6. Daniele Bielli, Christian Ferko, Michele Galli, Gabriele Tartaglino-Mazzucchelli, "Higher-spin currents and flows in auxiliary field sigma models", JHEP 08, 078 (2025), arXiv:2504.17294, DOI:10.1007/JHEP08(2025)078, https://arxiv.org/abs/2504.17294
  7. Christian Ferko, Michele Galli, Zejun Huang, Gabriele Tartaglino-Mazzucchelli, "Soliton Surfaces and the Geometry of Integrable Deformations of the CPN−1\mathbb{CP}^{N-1} Model", JHEP 03, 144 (2026), arXiv:2509.05081, DOI:10.1007/JHEP03(2026)144, https://arxiv.org/abs/2509.05081
  8. Philippe Lecheminant, Yuya Tanizaki, Keisuke Totsuka, "Infrared properties of two-dimensional SU(N)/H nonlinear sigma models at nonzero theta angles", SciPost Phys. 18, 183 (2025), arXiv:2412.17493, DOI:10.21468/SciPostPhys.18.6.183, https://arxiv.org/abs/2412.17493
  9. F. Delduc, T. Kameyama, S. Lacroix, M. Magro, B. Vicedo, "Ultralocal Lax connection for para-complex ZT\mathbb{Z}_T-cosets", Nucl. Phys. B 949, 114821 (2019), arXiv:1909.00742, DOI:10.1016/j.nuclphysb.2019.114821, https://arxiv.org/abs/1909.00742
  10. John Joseph M. Carrasco, Nicolas H. Pavao, "Even-point multi-loop unitarity and its applications: exponentiation, anomalies and evanescence", JHEP 01, 019 (2024), arXiv:2307.16812, DOI:10.1007/JHEP01(2024)019, https://arxiv.org/abs/2307.16812
  11. E. Abdalla, M. C. B. Abdalla, M. Gomes, "Nonlocal Charge of the CPN−1\mathbb{CP}^{N-1} Model and Its Supersymmetric Extension to All Orders", Phys. Rev. D 27, 825–836 (1983), DOI:10.1103/PhysRevD.27.825, https://inspirehep.net/literature/179294
  12. Benjamin Basso, Adam Rej, "On the integrability of two-dimensional models with U(1)xSU(N) symmetry", Nucl. Phys. B 866, 337–377 (2013), arXiv:1207.0413, DOI:10.1016/j.nuclphysb.2012.09.003, https://arxiv.org/abs/1207.0413
  13. Matthew Dodelson, Simeon Hellerman, Masataka Watanabe, Masahito Yamazaki, "Integrability of large-charge sectors in generic 2D EFTs", JHEP 08, 166 (2024), arXiv:2310.01823, DOI:10.1007/JHEP08(2024)166, https://arxiv.org/abs/2310.01823