Pith. sign in

REVIEW 2 major objections 6 minor 20 references

Lattice techniques to investigate the strong $CP$ problem: lessons from a toy model

T0 review · 2 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

Pith's one-line read A recent proposal to dissolve the strong CP problem by reordering lattice limits fails in the quantum rotor, the simplest theory with a θ term: exact results and simulations give the conventional nonzero topological susceptibility, not…

desk verdict Clean, internally sound rotor counterexample to a proposed order of limits, but the QCD relevance is unproven and the core result is already in the authors' earlier paper. read the letter →

arxiv 2502.01217 v2 pith:SIHFP25N submitted 2025-02-03 hep-lat hep-ph

classification hep-lathep-ph PACS 11.15.Ha11.30.Er02.70.Uu
keywords strongCPproblemquantumrotortopologicalsusceptibilityorderoflimitsthetatermlatticefieldtheorytopologyfreezingsign
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tests a recent proposal that the strong CP problem can be dissolved by a subtlety in how lattice limits are taken: first send the volume to infinity, and only afterwards sum over topological sectors. Applying that order of limits to the quantum rotor, the simplest theory with a θ term and nontrivial topology, the authors find that the topological susceptibility vanishes. The exact quantum-mechanical spectrum instead gives $\chi_t = 1/(4\pi^2 I)$, and with the conventional order of limits this is also the value that lattice simulations approach in the continuum limit. The paper concludes that the proposed no-CP-violation mechanism fails in the simplest topological model, and that the θ angle does affect physical observables. Along the way it presents two algorithmic devices, a winding hybrid Monte Carlo step and truncated-polynomial automatic differentiation, that overcome topology freezing and the sign problem in this model.

What carries the argument

The quantum rotor, a free particle of mass $m$ on a ring with Hamiltonian $H = -\frac{1}{2I}\left(\partial_\phi - \frac{\theta}{2\pi}\right)^2$ and moment of inertia $I = mR^2$, whose spectrum is exactly solvable and supplies the reference value $\chi_t = 1/(4\pi^2 I)$. The order-of-limits identity of Eq. (2) is the target of the test: with the Gaussian sector distribution $p_T(Q) \propto \exp(-2\pi^2 I Q^2/T)$, taking $T \to \infty$ before un-restricting the sector sum makes every term of order $1/T$ vanish. Two algorithmic devices carry the numerical argument: the winding transformation $W_\pm: \phi_t \to \phi_t \pm 2\pi t/\hat{T}$, an exact one-step change of the topological charge embedded into the HMC algorithm (wHMC) to restore ergodicity; and the algebra of truncated polynomials, applied either by reweighting the $\theta = 0$ ensemble or directly inside the HMC equations of motion (HAD), which extracts Taylor coefficients in θ from a single simulation without the noisy disconnected contributions of reweighting, subject to the caveat that the Metropolis accept-reject step is not differentiable and must be integrated precisely.

What would settle it

Simulate the quantum rotor following the proposed order of limits literally—restrict the ensemble to sectors |Q| < N at finite T, send T to infinity, then let N grow—and check whether the topological susceptibility vanishes; the paper predicts zero for every finite N, so any nonzero limit as N and T grow would show the refutation fails.

Watch

Extended reading notes

Core claim

The central claim is that the order of limits proposed in recent work—infinite volume taken before the sum over topological sectors, Eq. (2)—when applied to the quantum rotor produces a zero topological susceptibility, while the exact spectrum $E_n = \frac{1}{2I}\left(n - \frac{\theta}{2\pi}\right)^2$ gives $\chi_t = d^2E_0/d\theta^2|_{\theta=0} = 1/(4\pi^2 I)$, and lattice simulations agree with the conventional order of limits. Because the sector distribution $p_T(Q)$ is Gaussian with width proportional to $\sqrt{T}$, the proposed double limit is dominated by the $Q = 0$ sector and yields zero for any finite sector cutoff $N$. The simulations, run very close to the continuum thanks to the winding algorithm, extrapolate to the exact value, and they also confirm the linear θ-dependence of the first excited level, $\Delta E_1 = \frac{1}{2I}\left(1 - \frac{\theta}{\pi}\right)$. The paper therefore asserts that the proposal, which would make θ disappear from all observables and remove the strong CP problem without new physics, fails already in the simplest theory that shares the essential features of topology and a θ term.

Load-bearing premise

The refutation assumes that the order-of-limits proposal is meant to apply to the quantum rotor, identifying the rotor's Euclidean time extent T with the spacetime volume V of QCD; if the proposal relies on features specific to QCD, such as confinement or the volume scaling of gauge-field sectors, a counterexample in the rotor would not refute it for the real theory.

Editorial extensions

If this is right

  • If the rotor result is representative, the strong CP problem stands as conventionally formulated: the θ angle cannot be argued out of existence by a choice of limit order, so the smallness of θ still needs an explanation.
  • The winding HMC algorithm removes topology freezing in the rotor, with autocorrelation times of the topological charge saturating rather than growing exponentially toward the continuum, making the continuum extrapolation feasible.
  • Truncated-polynomial techniques (reweighting and HAD) recover the θ-expansion coefficients from a single simulation at θ = 0, with HAD reducing the statistical error by about an order of magnitude relative to direct fits.
  • Both the topological susceptibility and the first excited energy level of the rotor approach their exact quantum-mechanical values in the continuum limit, validating the conventional order of limits in this model.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If this counterexample transfers to QCD, then the recent no-CP-violation proposal is refuted and the strong CP problem still demands new physics such as an axion; the transfer is exactly the paper's load-bearing assumption, since QCD-specific features like confinement could in principle rescue the proposal.
  • The Gaussian-width argument suggests a general pattern: in any theory whose topological-charge distribution has width growing as the square root of the volume, the proposed order of limits will suppress the susceptibility to zero, so the rotor is likely not an isolated counterexample.
  • A decisive next test would be to implement the same order of limits in a two-dimensional model with known θ-dependence, such as CP(N-1), where the wHMC and HAD techniques could be ported; a vanishing susceptibility there would close the debate in the paper's favor.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 6 minor

Summary. The paper studies the quantum rotor with a theta term as a toy model for the strong CP problem. It solves the model exactly, obtains the topological-sector distribution, and evaluates the order-of-limits proposal of Refs. [6,7] (Eq. (2)): with the limits N and T taken in that order, the topological susceptibility vanishes, while the conventional thermodynamic limit yields chi_t = 1/(4 pi^2 I) (Eqs. (9)-(10)). The paper then introduces a winding HMC algorithm to avoid topology freezing and uses truncated polynomials / Hamiltonian Automatic Differentiation (HAD) to compute theta-derivatives without the sign problem. Lattice results for a local susceptibility and the first theta-dependent energy gap agree with the analytic conventional result. The final section states that the toy-model results disagree with the claims of Refs. [6,7].

Significance. The exact rotor solution is clean and the proposed order of limits is evaluated correctly; no fitted parameter enters the central comparison, and the algorithmic techniques are benchmarked against exact analytic results. If the order-of-limits proposal of Refs. [6,7] is taken to apply to this system, the paper provides a sharp counterexample. The transferability of that proposal from QCD to the rotor is, however, an extra premise, so the significance is that of a controlled toy-model lesson rather than a direct QCD disproof.

major comments (2)
  1. [Sec. 1, Eq. (2), Sec. 5] The paper's central negative claim about Refs. [6,7] rests on the premise, stated in Sec. 1, that 'the claims presented in Refs. [6,7] should also hold in simpler models.' This premise is not argued for. In the rotor, T is the full spacetime volume, so the limits lim_{T->infty} and lim_{V->infty} in Eq. (9) coincide, and there is no separate spatial volume in which the distribution over topological sectors could behave differently from the one-dimensional Euclidean-time case. The final Sec. 5 sentence ('disagree with the claims of Refs. [6,7]') is therefore stronger than what the model establishes. Please either justify the transfer (for instance by deriving Eq. (2) directly from the rotor path integral and by explaining why QCD-specific features such as a 4D sector distribution or confinement are not needed) or explicitly restrict the conclusion to the toy model. A concrete way to test the transferability would be to repeat the analysis in a theory with separate spatial and temporal extents, such as a 1+1D sigma model with a topological term.
  2. [Sec. 5, Fig. 5 (left)] The text says the simulations 'validate the conventional order of limits,' but the plotted observable chi_t = <(phi_1 - phi_0)^2> is a local correlator that is independent of the topological-sector sum; it cannot distinguish the two orders of limits. The analytic result in Eq. (10) is what establishes the conventional-order value, and the lattice data confirm that the discretized action reproduces the analytic continuum value. To support the validation sentence, present the sector-summed susceptibility (e.g., <Q^2>_T/T with the wHMC ensembles) or rephrase the claim as a check of the lattice formulation against the analytic conventional result.
minor comments (6)
  1. [Sec. 4, first paragraph] The word 'analiticity' should be 'analyticity'.
  2. [Eq. (13)] The notation W+ W- = I is unclear: W+ and W- are maps on configurations rather than operators, and the identity map should be defined explicitly.
  3. [Fig. 5] Please describe the continuum extrapolation procedure quantitatively (fit form, fitted range in 1/I-hat, and goodness of fit) rather than only stating agreement with the dashed analytic lines.
  4. [Footnote 2] Since HAD omits the Metropolis accept-reject step, please report a numerical check of the integration precision (e.g., dependence on the molecular-dynamics step size) to quantify the acknowledged systematic uncertainty.
  5. [Sec. 2, Eqs. (2) and (9)] The switch from V in Eq. (2) to T in Eq. (9) should be stated explicitly as V = T for the rotor, to avoid the impression that a spatial volume and the inverse temperature are being conflated.
  6. [References] References [15] and [19] are listed as unpublished; if published versions exist, they should be cited.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the rotor exact results are derived in-paper and the self-citations are non-load-bearing.

full rationale

The derivation chain is self-contained. The rotor spectrum Eq. (6) is obtained directly from the Hamiltonian Eq. (3), and the topological-charge distribution p_T(Q) in Eq. (7) follows from the partition function by Fourier transform. The proposed-order result in Eq. (9) is a direct evaluation of Eq. (2) using that p_T(Q): for fixed N the ratio [sum Q^2 exp(-2 pi^2 I Q^2/T)]/[sum exp(-2 pi^2 I Q^2/T)] is O(N^2) before division by T, so the T->infinity limit is zero and the subsequent N->infinity limit does not change it. The conventional result Eq. (10) is the second derivative of the same exact spectrum. No parameter is fitted to make these two evaluations agree; they disagree by direct calculus. The lattice simulations in Fig. 5 are benchmarked against independent analytic results with open boundary conditions and are not used to tune the central comparison. The self-citations [14,18,20] concern a previously published version of the same rotor calculation, the wHMC algorithm (which is also defined in Eqs. (13)-(14)), and a software library; none carries a load-bearing premise that is unverified in this paper. The only substantive caveat is the transferability assumption stated in Sec. 1—that the proposal of Refs. [6,7] should also hold in simpler models—but that is a scope/domain assumption, not a circular derivation. Hence no circularity.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No free parameters are fitted; all numerical results are compared to the exact rotor solution. The main imported premise is that Refs. [6,7]'s order-of-limits proposal should hold in the rotor. Additional domain assumptions concern the lattice discretizations and analytic continuation in theta. No invented physical entities are introduced.

assumptions (4)
  • domain assumption The order-of-limits proposal of Refs. [6,7] should apply to the quantum rotor.
    Sec. 1 states the claims of Refs. [6,7] should also hold in simpler models; the whole counterexample depends on this transferability.
  • domain assumption The Euclidean time extent T of the rotor plays the role of spacetime volume V in Eq. (2).
    Sec. 2 applies the proposed order of limits to T as the volume, identifying the rotor's finite-temperature extent with the QCD volume.
  • domain assumption The standard and classical perfect lattice discretizations have the same continuum limit at fixed T/I.
    Sec. 3 asserts both discretizations lead to the continuum limit of Eq. (5) along lines of constant physics.
  • domain assumption Analytic continuation in the imaginary theta parameter theta_I is valid for the observables considered.
    Sec. 4 relies on analyticity of observables O(theta) to extract Taylor coefficients from simulations at imaginary theta.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Lattice techniques to investigate the strong $CP$ problem: lessons from a toy model." pith.science (2026). https://pith.science/paper/SIHFP25N

@misc{pith2026250201217,
  author       = {Pith},
  title        = {Pith review of: Lattice techniques to investigate the strong $CP$ problem: lessons from a toy model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SIHFP25N}},
  note         = {Machine review of arXiv:2502.01217}
}
abstract

Recent studies have claimed that the strong $CP$ problem does not occur in QCD, proposing a new order of limits in volume and topological sectors when studying observables on the lattice. We study the effect of the topological term on a simple quantum mechanical rotor that allows a lattice description. We particularly focus on recent proposals to face the challenging problems that this study poses in lattice QCD and that are also present in the quantum rotor, such as topology freezing and the sign problem.

Figures

Figures reproduced from arXiv: 2502.01217 by the authors.

Figure 1
Figure 1. (Left) Representation of the quantum rotor. (Middle and right) Possible trajectories of the quantum rotor along a torus in the path integral representation, with topological charges 𝑄 = 0 and 𝑄 = 1. The system can also be formulated as a path integral at finite temperature 𝛽 and Euclidean volume 𝑇 = 1/𝛽 via the partition function [15, 16] 𝑍𝑇 (𝜃) = ∫ D𝜙 𝑒−𝑆[𝜙]+𝑖 𝜃𝑄[𝜙] , (4) where action and topological charge in the … view at source ↗
Figure 2
Figure 2. Winding transformation on a quantum rotor configuration on the lattice. which is trivially zero. Alternatively, one can obtain the topological susceptibility in the zero temperature limit (𝑇 → ∞) directly from the energy spectrum in Eq. (6), which reads 𝜒𝑡 ≡ 𝑑 2𝐸0 (𝜃) 𝑑𝜃2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Autocorrelation time of 𝑄 as a function of ˆ𝐼 for wHMC (blue) and HMC (orange), keeping 𝑇ˆ/ ˆ𝐼 = 100, with the standard discretization of the action. The simplest idea to build an algorithm that can sample topology efficiently is to find transfor￾mations between topological sectors. Such a topology-changing transformation can be easily built for the quantum rotor, for which we define the winding transformation by W±… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: The 𝜃𝐼 dependence of 𝐼 𝜒𝑡 at ˆ𝐼 = 5 and 𝑇ˆ = 100 computed with different methods for periodic boundary conditions and the standard definitions of the action and topological charge. (Left): results (orange circles) from five different simulations with 100k uncorrelated …
Figure 5
Figure 5. Figure 5: (Left) Continuum extrapolation of 𝜒𝑡 computed from ⟨(𝜙1 − 𝜙0) 2 ⟩ at various {𝑇,ˆ ˆ𝐼} at constant 𝑇/𝐼 = 20 with the standard action and classical topological charge (red, filled circles), with the classical action and topological charge (green, open circles) and from m…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 9 canonical work pages

  1. [14]

    Albandea, G

    D. Albandea, G. Catumba and A. Ramos,Strong CPproblem in the quantum rotor,Phys. Rev. D110 (2024) 094512 [2402.17518]

  2. [1]

    Abel et al.,Measurement of the Permanent Electric Dipole Moment of the Neutron,Phys

    C. Abel et al.,Measurement of the Permanent Electric Dipole Moment of the Neutron,Phys. Rev. Lett.124 (2020) 081803 [2001.11966]

  3. [2]

    Chupp, P

    T. Chupp, P. Fierlinger, M. Ramsey-Musolf and J. Singh,Electric dipole moments of atoms, molecules, nuclei, and particles, Rev. Mod. Phys.91 (2019) 015001 [1710.02504]

  4. [3]

    Peccei and H.R

    R.D. Peccei and H.R. Quinn,Constraints imposed byCPconservation in the presence of pseudoparticles,Phys. Rev. D16 (1977) 1791

  5. [4]

    Nelson,Naturally WeakCPViolation, Phys

    A.E. Nelson,Naturally WeakCPViolation, Phys. Lett. B136 (1984) 387

  6. [5]

    Barr,Solving the strongCPproblem without the Peccei-Quinn symmetry, Phys

    S.M. Barr,Solving the strongCPproblem without the Peccei-Quinn symmetry, Phys. Rev. Lett. 53(1984) 329

  7. [6]

    W.-Y. Ai, J.S. Cruz, B. Garbrecht and C. Tamarit,Absence of CP violation in the strong interactions, 2001.07152

  8. [7]

    W.-Y. Ai, J.S. Cruz, B. Garbrecht and C. Tamarit,Consequences of the order of the limit of infinite spacetime volume and the sum over topological sectors for CP violation in the strong interactions, Phys. Lett. B822(2021) 136616

Show all 20 references
  1. [8]

    de Forcrand,Simulating QCD at finite density, PoS LAT2009(2009) 010 [1005.0539]

    P. de Forcrand,Simulating QCD at finite density, PoS LAT2009(2009) 010 [1005.0539]

  2. [9]

    Gattringer and K

    C. Gattringer and K. Langfeld,Approaches to the sign problem in lattice field theory,Int. J. Mod. Phys. A31 (2016) 1643007 [1603.09517]

  3. [10]

    Alles, G

    B. Alles, G. Boyd, M. D’Elia, A. Di Giacomo and E. Vicari,Hybrid Monte Carlo and topological modes of full QCD,Phys. Lett. B389 (1996) 107 [hep-lat/9607049]

  4. [11]

    Del Debbio, H

    L. Del Debbio, H. Panagopoulos and E. Vicari,Theta dependence of SU(N) gauge theories, JHEP 08(2002) 044 [hep-th/0204125]

  5. [12]

    Del Debbio, G.M

    L. Del Debbio, G.M. Manca and E. Vicari,Critical slowing down of topological modes, Phys. Lett. B594 (2004) 315 [hep-lat/0403001]

  6. [13]

    ALPHA collaboration,Critical slowing down and error analysis in lattice QCD simulations, Nucl. Phys. B845(2011) 93 [1009.5228]

  7. [15]

    Fjeldso, J

    N. Fjeldso, J. Midtdal and F. Ravndal,Random walks of a quantum particle on a circle, unpublished (1987)

  8. [16]

    Bietenholz, R

    W. Bietenholz, R. Brower, S. Chandrasekharan and U.J. Wiese,Perfect lattice topology: The Quantum rotor as a test case,Phys. Lett. B407 (1997) 283 [hep-lat/9704015]. 9 Lattice techniques to investigate the strong𝐶𝑃 problem: lessons from a toy model David Albandea

  9. [17]

    Duane, A.D

    S. Duane, A.D. Kennedy, B.J. Pendleton and D. Roweth,Hybrid Monte Carlo,Phys. Lett. B 195 (1987) 216

  10. [18]

    Albandea, P

    D. Albandea, P. Hernández, A. Ramos and F. Romero-López,Topological sampling through windings,Eur. Phys. J. C81(2021) 873 [2106.14234]

  11. [19]

    Haro,Automatic differentiation tools in computational dynamical systems, unpublished (2011)

    A. Haro,Automatic differentiation tools in computational dynamical systems, unpublished (2011)

  12. [20]

    Ramos,FormalSeries.jl, May, 2023

    A. Ramos,FormalSeries.jl, May, 2023. 10.5281/zenodo.7970278. 10

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.