Quantum-Classical Correspondence of Non-Hermitian Symmetry Breaking

Hai-Dong Li; Wei Chen; Zhuo-Ting Cai

Quantum-Classical Correspondence of Non-Hermitian Symmetry Breaking

T0 review · 3 major / 3 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read In η-pseudo-Hermitian systems, whether each energy level is real or complex is decided by the Sη symmetry class of the periodic orbit that quantizes it.

desk verdict A serious semiclassical framework for non-Hermitian spectra, but the central symmetry equation as printed is false for the paper's own first example, so the universal-correspondence claim needs repair before it is established. read the letter →

arxiv 2411.17398 v2 pith:WPWVZULJ submitted 2024-11-26 quant-ph cond-mat.mes-hallmath-phmath.MP

Zhuo-Ting Cai , Hai-Dong Li , Wei Chen This is my paper · ORCID

classification quant-phcond-mat.mes-hallmath-phmath.MP MSC 81Q1281Q5081S40 PACS 03.65.Sq11.30.Er

keywords non-Hermitianphysicspseudo-Hermitiansymmetryparity-timeGutzwillertraceformulasemiclassicalorbitscomplexpathintegralsexceptionalpointsquantum-classicalcorrespondence

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 claims a universal per-state rule for non-Hermitian spectral transitions: in any analytic system with η-pseudo-Hermitian symmetry, each energy level is tied to periodic orbits in complex phase space, and the symmetry class of the orbit dictates whether the level is real or complex. Real levels correspond to orbits that are invariant under the classical Sη symmetry; complex-conjugate pairs correspond to two orbits that individually break Sη but are mapped into each other by it. The rule is established by extending the Gutzwiller trace formula to non-Hermitian systems through complex path integrals, then checking it against numerical spectra of skin-effect, nonreciprocal-lattice, double-well, and two-level PT models. If correct, it provides a physical mechanism for non-Hermitian symmetry breaking at the level of individual states and explains why the exceptional point is inherently quantum, with no single classical orbit describing it.

What carries the argument

The load-bearing object is the generalized Gutzwiller trace formula in complex phase space, G(E)=iT(E)$e^{{i(∮p dx−2πμ)}}$/(1−$e^{{i(∮p dx−2πμ)}}$), whose poles yield the quantization condition ∮O p dx = (n+μ)2π. Around this sits the classical Sη symmetry H(xη,pη)=H*(x*,p*), which maps periodic orbits into Sη-symmetric pairs and, through the paper's theorems on contour integrals, energies, and periods, forces each quantized level to be either real (self-symmetric orbit) or half of a complex-conjugate pair (two paired orbits). A third ingredient is the spin coherent-state path integral that adapts the same machinery to two-level systems.

What would settle it

Compute the exact spectrum of any η-pseudo-Hermitian model with known analytic eigenfunctions, extract the complex periodic orbits at each eigenenergy, and evaluate the orbit-symmetry difference measure F from Eq. (S.140) of the Supplemental Material: one real level whose orbit has F>0, or one complex pair whose two orbits have F=0, would refute the central dichotomy.

Watch

Extended reading notes

Core claim

The paper asserts that for any analytic non-Hermitian Hamiltonian obeying η-pseudo-Hermiticity, the quantum η symmetry has an exact classical counterpart Sη given by H(xη,pη)=H*(x*,p*). By applying saddle-point approximations to complex path integrals, the authors derive a generalized Gutzwiller trace formula whose poles give the quantization condition ∮O p dx = (n+μ)2π along periodic orbits in complex phase space. They then prove that Sη forces exactly two orbit configurations: an orbit invariant under Sη, whose quantized energy is real, or an Sη-symmetric pair of distinct orbits, whose quantized energies are complex conjugates. This dichotomy is verified numerically for four models: a non-Hermitian skin-effect Hamiltonian, a nonreciprocal lattice in a magnetic field, a PT-symmetric double well, and a PT-symmetric two-level system. Near the exceptional point, single-orbit quantization fails and quantum tunneling between orbits—or, in the two-level case, divergence and abrupt reorientation of the classical spin—is required, which the paper presents as evidence that the exceptional point is intrinsically quantum.

Load-bearing premise

The derivation assumes that equality of the two complex path integrals for all endpoints and times forces equality of their integrands, yielding the classical Sη symmetry; for path integrals with nontrivial measures this implication is a semiclassical leap, stated in the Supplemental Material around Eqs. (S.13)-(S.14), rather than a proven theorem.

Editorial extensions

If this is right

The real-to-complex transition of every non-Hermitian level can be tracked to a symmetry change of a single semiclassical orbit, making spectral phase transitions per-state phenomena rather than only global ones.
The generalized trace formula gives a practical quantization rule: find complex periodic orbits of the classical equations of motion, apply ∮p dx=(n+μ)2π, and the resulting energies match numerical spectra across the models tested.
In continuous models, the exceptional point appears as the breakdown of single-orbit quantization; including tunneling between orbits through the generalized condition restores agreement with exact numerics, marking the exceptional point as a quantum effect.
For PT-symmetric two-level systems, the PT transition is interpreted geometrically: real spectra correspond to classical spin orbits about ±M with SPT symmetry, while complex spectra correspond to paired orbits about ±iM that individually break the symmetry.
Because the framework covers both continuous and discrete systems, the same orbit-symmetry labeling applies to lattice models and few-level non-Hermitian systems.
If the per-orbit labeling is exact, designing a potential or gain-loss profile that forces an Sη symmetry change on one orbit should switch that level's reality without changing others, suggesting orbit-selective control of non-Hermitian spectra.
The tunneling correction near exceptional points suggests that in higher-dimensional or many-body non-Hermitian systems, exceptional points may generically appear where two families of periodic orbits merge, potentially connecting the result to level-avoidance statistics in complex spectra.
Because the argument relies only on pseudo-Hermiticity and analyticity, the same orbit-symmetry dichotomy should hold for other discrete symmetry classes that can be cast as η-PHS, a testable extension beyond PT and MT symmetries.

Reading between the lines

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

If the per-orbit labeling is exact, designing a potential or gain-loss profile that forces an Sη symmetry change on one orbit should switch that level's reality without changing others, suggesting orbit-selective control of non-Hermitian spectra.
The tunneling correction near exceptional points suggests that in higher-dimensional or many-body non-Hermitian systems, exceptional points may generically appear where two families of periodic orbits merge, potentially connecting the result to level-avoidance statistics in complex spectra.
Because the argument relies only on pseudo-Hermiticity and analyticity, the same orbit-symmetry dichotomy should hold for other discrete symmetry classes that can be cast as η-PHS, a testable extension beyond PT and MT symmetries.

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.

Desk Editor's Note

read the letter

Two things to know before you read it. It is a serious attempt at the open problem named in the abstract: attaching a semiclassical orbit to each individual non-Hermitian eigenstate and predicting, from the orbit's symmetry, whether the energy is real. The generalized Gutzwiller trace formula and the orbit-pairing picture are genuinely new, and the numerical checks in the four examples are convincing at the level of the quantization conditions. The treatment of the exceptional point as a quantum phenomenon, with tunneling corrections, is also a real contribution.

The second thing is a problem. The paper's central displayed symmetry, Eq. (4) (and SM Eq. (S.14)), is not satisfied by the paper's first example. For H1 = (p+iγ)^2 + V0|x|, the printed ST condition H1(x,-p) = H1^*(x*,p*) fails at the symmetric-orbit point x=0, p=√E−iγ: the left side is (−√E+2iγ)^2, the right side is E. The orbit pairing O2(t*) = {x^*(-t), -p^*(-t)} that the example actually uses follows from H1(x,-p) = H1^*(x,p) — complex conjugation of the coefficients with the arguments unstarred. The SM Sec. II derivation reaches the starred form only after renaming the conjugated integration variables; carrying that renaming through consistently gives H^*(x,p), not H^*(x^*,p^*). So the displayed equation is wrong, and the theorem proofs in Sec. IV rest on it.

I am not calling the whole project a wash. The intended statement is clear, and the correspondence likely survives a fix: with the corrected symmetry, the orbit-pairing theorems should go through with modest changes. But as printed, the universal claim is not established. A second soft spot is the leap from equality of path integrals to equality of integrands; that is a genuine gap. The Maslov indices are assigned by hand, and the EP tunneling correction is an add-on that works in the double-well example but is not derived from the trace formula.

Bottom line: this paper deserves a serious referee, but it will need major revision. The numerical work is strong, and the framework is worth engaging. Fix the symmetry equation, re-derive the orbit-pairing theorem, and either prove or soften the integrand-equality step. If those corrections land, it becomes a useful reference; for now, treat the universal claim with caution.

Referee Report

3 major / 3 minor

Summary. The paper develops a quantum-classical correspondence for non-Hermitian systems with eta-pseudo-Hermitian symmetry. Using a complex path integral formulation and a generalized Gutzwiller trace formula, it claims that real eigenenergies are quantized along periodic orbits preserving the classical S_eta symmetry, while complex conjugate eigenenergies arise from orbits that individually break S_eta but are exchanged by it. The claim is supported by four examples: a skin-effect Hamiltonian H1, a nonreciprocal lattice with magnetic field H2, a PT-symmetric double-well H3, and a PT-symmetric two-level system H4. The Supplemental Material contains the derivations of the propagator constraint, the S_eta symmetry, the non-Hermitian Gutzwiller trace formula, the orbit-symmetry theorems, detailed calculations for the examples, and a tunneling-based treatment of exceptional points.

Significance. If correct, the proposed correspondence would be an important structural result: it attaches a concrete semiclassical label to every non-Hermitian eigenenergy and explains, at the level of individual states, why some eigenvalues are real and others occur in complex conjugate pairs. The paper has notable strengths: the quantization conditions are used without fitting continuous parameters, they are compared directly against numerical diagonalization in several models, and the central classification into real versus paired-complex energies is testable through orbit geometry. The demonstration that exceptional points require quantum tunneling corrections, rather than a single classical orbit, is also a substantive and falsifiable claim. However, the central derivation contains a load-bearing symmetry-error and an unjustified path-integral inference, as detailed below.

major comments (3)

[Eq. (4) and SM Sec. II, Eq. (S.14)]
[Eq. (3) and SM Sec. II around Eq. (S.13)]
[SM Sec. III, Eqs. (S.15)-(S.23)]

minor comments (3)

[Eq. (1) and throughout]
[Supplemental Material, global]
[Main text, Example 1 paragraph]

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the real/complex energy classification is derived from the Sη orbit symmetry and the quantization condition, with spectra checked against independent numerical diagonalization rather than fitted.

full rationale

The central derivation is self-contained against external benchmarks. The step from the complex path-integral equality, Eq. (3), to the classical Sη symmetry, Eq. (4), is an inference from equality of path integrals for arbitrary endpoints and times to equality of integrands; this is a semiclassical assumption and a possible source of error, but it is not a circular step because it does not presuppose the real/complex classification it is used to explain. The orbit-pairing theorems in Supplemental Material Sec. IV derive the action-conjugation identity and E1 = E2* from the Sη symmetry of the classical Hamiltonian, and the distinction between real energies and complex-conjugate pairs follows from whether the paired orbits coincide; this distinction is not inserted by definition. In the worked examples, the quantization conditions are solved for energies and then compared with numerical solutions of the eigenvalue equations; no continuous parameter is fitted to the target spectrum. Self-citations such as Ref. [42] provide background (a model and the notion of complex time) but are not load-bearing premises of the trace-formula derivation, which is rederived in the Supplemental Material. A possible algebraic issue with the placement of complex-conjugation stars in Eqs. (S.13)-(S.14) would be a correctness defect, not circularity, because it does not make the predictions equal to the inputs by construction. Therefore no circular step is exhibited, and the circularity score is 0.

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

No continuous parameters are fitted to the spectra. Maslov indices (1/2 or 0) are assigned based on the number of turning points in each orbit and are not adjusted to match numerics. No new physical entities are introduced; complex periodic orbits are mathematical constructs emerging from the complex path integral.

assumptions (4)

domain assumption The complex path integral for the propagator with complex coordinates, momenta, and time is a valid representation of the non-Hermitian quantum dynamics.
Used at the start of SM Sec. II and Eq. (5) of the main text; the analytic continuation of the path integral is assumed to preserve the spectral information.
ad hoc to paper Equality of complex path integrals for arbitrary endpoints and times forces equality of the classical integrands, leading to H(xη,pη)=H*(x*,p*).
Inferred in SM Sec. II from Eq. (S.13) to Eq. (S.14). This is not generally true for path integrals with nontrivial measures; it is a strong semiclassical assumption.
domain assumption The η similarity transformation induces an extended canonical transformation in phase space satisfying M J M^T = ∓J.
Proved in SM Sec. IV using [xη,pη]=∓i; the sign depends on whether η involves a transpose. This is assumed for all pseudo-Hermitian symmetries considered.
domain assumption The saddle point approximation applied successively to the complex path integral, the spatial trace, and the time integral yields the trace formula (Eq. 6) with no additional corrections beyond Maslov indices.
SM Sec. III; the SPA is assumed to be applicable in the complex domain and to give the exact pole structure of G(E).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantum-Classical Correspondence of Non-Hermitian Symmetry Breaking." pith.science (2026). https://pith.science/paper/WPWVZULJ

@misc{pith2026241117398,
  author       = {Pith},
  title        = {Pith review of: Quantum-Classical Correspondence of Non-Hermitian Symmetry Breaking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WPWVZULJ}},
  note         = {Machine review of arXiv:2411.17398}
}

abstract

Real-to-complex spectral transitions and the associated spontaneous symmetry breaking of eigenstates are central to non-Hermitian physics, yet a comprehensive and universal theory that precisely describes the underlying physical mechanisms for each individual state remains elusive. Here, we resolve the mystery by employing the complex path integral formalism and developing a generalized Gutzwiller trace formula. These methodologies enable us to establish a universal quantum-classical correspondence that precisely links the real or complex nature of individual energy levels to the symmetry properties of their corresponding semiclassical orbits. Specifically, in systems with a general $\eta$-pseudo-Hermitian symmetry, real energy levels are quantized along periodic orbits that preserve the corresponding classical $S_\eta$ symmetry. In contrast, complex conjugate energy levels arise from semiclassical orbits that individually break the $S_\eta$ symmetry but together form $S_\eta$-symmetric pairs. This framework provides a unified explanation for the spectral behaviors in various continuous non-Hermitian models and for the $\mathcal{PT}$ transition in two-level systems. Besides, we demonstrate that the exceptional point is inherently a quantum phenomenon, as it cannot be described by a single classical orbit. Our work uncovers the physical mechanism of non-Hermitian symmetry breaking and introduces a new perspective with broad implications for the control and application of non-Hermitian phenomena.

Discussion (0). Continue with ORCID to comment.

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.