{"id":"126aae74-7526-4f8a-b9ce-1927d81415b0","arxiv_id":"1908.06383","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In a PT-symmetric waveguide, increasing the distance between gain and loss generates a ladder of resonances that can be converted into a ladder of complex eigenvalues by tuning the gain-loss amplitude.","lead":"This paper studies a one-dimensional waveguide with one amplifying and one absorbing segment separated by a variable gap. It shows that widening the gap creates new resonances and spectral singularities and can lower the gain needed for laser-absorber operation.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Non-uniformity of Lemma 4's remainder leaves the ladder-sign claim not fully proven.","rationale":"The paper's rigorous lemmas and explicit characterizations are valuable, and the conditional verdict already reflects the main gap. My stress-test identifies the same weakest assumption as the Reader: the asymptotic expansion in Lemma 4 is not proved uniformly for n of order ℓ, and that uniformity is exactly what is needed for the headline claim that every member of the ladder has imaginary part sign given by sin√(2γ). The omitted 'delicate analysis' for sin√(2γ)=0 is an additional, but secondary, gap. I do not see a reason to reject the paper; the central construction may be correct, but the ladder-sign conclusion should be stated conditionally on a uniform remainder or explicitly weakened to a subsequence statement until the uniformity is supplied.","tokens_in":625,"tokens_out":4504,"duration_ms":136028,"concrete_test":"Take γ with sin√(2γ)>0 (e.g. γ=1) and γ with sin√(2γ)<0 (e.g. γ=10), choose a large ℓ so that (3.9) admits n = floor(2Θ(γ)ℓ/π - 1/2), and compute kn by high-precision numerical root-finding of (2.9) near πn/(2ℓ). Compare Im kn with the leading ℓ^-3 term of (3.10); any sign disagreement for an admissible n would falsify the ladder-sign claim. Analytically, re-derive (3.10) retaining n as an explicit parameter and establish sup_{n satisfying (3.9)} |ℓ^4(kn - four-term expansion)| < ∞ as ℓ→∞.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weakest link is Lemma 4, whose asymptotic expansion (3.10), with remainder O(ℓ^-4), is used in §4.2 to assert that every zero kn in the ladder has imaginary part whose sign is determined by sin√(2γ). The problem is that Lemma 4 allows n to grow with ℓ: condition (3.9) permits |n| as large as (2Θ(γ)/π)ℓ, i.e. n of order ℓ. In the proof of (3.10), equation (3.16) is expanded near ξ = πn/2 with ε = ℓ^-1. For fixed n this is a legitimate Taylor expansion, but for n ~ ℓ the point πn/2 is itself O(ε^-1), and no uniform bound for the O(ε^3) remainder is proved. If the remainder grows with n as, say, n^4, then for n ~ ℓ the error term is O(1) rather than O(ℓ^-3), and the leading imaginary part in (3.10), which is only O(ℓ^-1) for the outermost zeros, could be overwhelmed. The claim in §4.2 is not merely that some subsequence has this sign, but that the entire ladder of O(ℓ) zeros lies above or below the real axis according to sin√(2γ); that conclusion requires uniformity over the full range (3.9). The paper also flags, for sin√(2γ)=0, a 'delicate analysis' showing Im k = O(ℓ^-5) without supplying that analysis; even if the uniform remainder were established, this exceptional case would still need a proof or an explicit exemption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyses a one-dimensional Schrödinger operator (2.1) with a PT-symmetric potential formed by two separated gain/loss elements, reducing the spectral problem to the scalar transcendental equation F(k,ℓ,γ)=0 in (2.9). It establishes, via a sequence of lemmata, that F is entire with countably many zeroes accumulating only at infinity, that all but finitely many zeroes lie in an explicit sector, and that for large separation ℓ there are O(ℓ) simple zeroes near the real segment [-Θ(γ), Θ(γ)], with a four-term asymptotic expansion (Lemma 4). It also studies zeroes of the separated one-step problems (Lemmata 5–6), proves the absence of pure imaginary eigenvalues (Lemma 7), and derives a forbidden gap with no spectral singularities (Lemma 11). Section 5 develops an algorithm for spectral singularities at prescribed wavenumbers and uses numerical continuation to describe the PT-breaking threshold as a function of ℓ and branches of spectral singularities. The authors interpret the large-ℓ ladder as a Wannier–Stark-like phenomenon, with the sign of the imaginary parts switchable between resonances and eigenvalues by varying sin√(2γ).","tokens_in":27919,"tokens_out":10625,"duration_ms":102359,"significance":"The paper is valuable as a simple, exactly tractable model in which spacing gain and loss produces qualitatively new spectral phenomena: a growing ladder of resonances or eigenvalues accumulating near an explicitly known real segment, and a mechanism for converting one kind of object into the other through the gain/loss amplitude. The derivation of (2.9) is clean, the asymptotic coefficients in (3.10) are computed rather than fitted, and the Rouché-based counting statements in Lemmata 4 and 6 are checkable and, in principle, machine-verifiable. The forbidden-gap statement (5.9) is crisp and sharp. The physical interpretation in terms of laser-absorber modes and PT-breaking is well grounded in the existing literature. The main caveat is that several headline claims—the full-ladder sign, the 'any wavenumber' spectral singularity, the monotone decrease of the threshold, and the O(ℓ^{-5}) case—are either not proved with uniformity or rest on numerical evidence.","major_comments":[{"comment":"Condition (3.9) permits |n| up to (2Θ(γ)/π)ℓ, i.e. n=O(ℓ), but the proof of (3.10) fixes n and expands (3.16) in ε=ℓ^{-1} around ξ=πn/2; no uniform bound for the remainder in that expansion over the full range of n is given. This matters because in §4.2 the sign of Im k_n for the whole ladder is read from the ℓ^{-3} term, which for n~ℓ is O(ℓ^{-1}), while a non-uniform remainder of order n^3 ε^3 or worse could be O(1) and would overwhelm the leading imaginary part. Please either prove uniformity over the range (3.9) using the explicit estimates already present in the proof, or restrict the sign/accumulation statement to a smaller range such as |n|≤Cℓ^{1-δ}, or reformulate the conclusion so that only existence and the four-term expansion for fixed n are claimed.","section":"§3, Lemma 4 / Eq. (3.10)"},{"comment":"The abstract and conclusion claim that a spectral singularity can be created at any wavenumber, but the only support is numerical: equation (5.11) is solved by dichotomy for k∈(0,10] with step Δk=0.01, and the admissible distances are then read from (5.3)–(5.4). No theorem states that for every k>0 there exists a suitable root β of (5.11) together with an integer n satisfying the sign condition (5.4). Please either prove existence, or qualify the claim as numerically demonstrated on a finite interval rather than as a proven result.","section":"§5.3, Eqs. (5.11), (5.3)–(5.4)"},{"comment":"The assertion that the PT-breaking threshold decreases monotonically with ℓ, and hence that spaced gain/loss lowers the laser-absorber threshold, is supported only by numerical continuation from ℓ=0; no analytical bound or error estimate is supplied. Since this is highlighted as a practical consequence, either provide a proof for a stated parameter range or explicitly label the monotonicity as numerical evidence.","section":"§5.4 / Fig. 5(a)"},{"comment":"The sentence 'A delicate analysis shows that in this case the imaginary parts … amount to O(ℓ^{-5})' is asserted without proof or reference. This case is exactly where the leading imaginary part in (3.10) vanishes, so the assertion is needed for the claimed ladder of 'nearly spectral singularities' at sin√(2γ)=0. Please include the analysis or state the case as an open problem.","section":"§4.2, paragraph after (3.10)"}],"minor_comments":[{"comment":"The text says the circle containing zeroes with Im k≤0 grows 'proportionally to γ', but from (3.6) r=max{39/20,√γ/2}, so for large γ the radius grows like √γ, not γ.","section":"§4.1, near (3.6)"},{"comment":"In the display after (3.8), the notation |Q1(z,ℓ,μ)| should read |Q1(z,μ)|, since Q1 depends only on z and μ.","section":"§3, proof of Lemma 3"},{"comment":"The sentence 'the potentials V± were introduced in (3.17)' is confusing: (3.17) is the change of variables in Lemma 5, while the potentials V± are defined in the displayed formula immediately after Lemma 4.","section":"§4.2"},{"comment":"There are several typographical slips that should be corrected in a revision: 'Schrödiner' in the abstract, 'contionuous' in §2.1, and 'we use consider' in §5.3.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper is in good shape analytically for the local statements, and I would not object to its eventual publication after revision. My main editorial concern is that the abstract and conclusion present as theorems several statements that are numerical or conditional: the 'any wavenumber' spectral singularity, the monotone threshold decrease, and the full-ladder sign for n~ℓ. The authors should either add the missing uniform estimates and the O(ℓ^{-5}) analysis, or carefully qualify these claims. The self-citations to [21,23,24] are appropriate background and do not raise a novelty concern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real news is that spacing gain and loss in a simple PT-symmetric step potential produces a Wannier-Stark-like ladder of resonances or eigenvalues near a segment of the real axis, something with no Hermitian counterpart. The paper derives a compact spectral equation, proves via Rouch\\'e that O(ℓ) zeros exist near the real axis for large ℓ, and computes an explicit four-term asymptotic expansion for them. That is genuine, mostly correct work.\n\nThe best part is Lemma 4: for each n satisfying the counting condition, a circle around πn/(2ℓ) contains exactly one zero, and the expansion coefficients are written in closed form. The authors also correctly stress how this differs from a true Wannier-Stark ladder, since here eigenvalues can emerge from ordinary points of the essential spectrum. Good.\n\nSoft spots, in order of importance. First, the uniformity gap in Lemma 4. The expansion (3.10) is derived as a Taylor expansion in ε = ℓ^{-1} with n held fixed, but condition (3.9) allows n ~ ℓ. For those outermost zeros the point ξ = πn/2 is itself O(ε^{-1}), and the remainder O(ℓ^{-4}) is not shown uniform in n. The paper then reads off the sign of Im k_n for every zero in the ladder from the ℓ^{-3} term. For n ~ ℓ that term is only O(ℓ^{-1}), so a non-uniform remainder could swamp it. The ladder existence is not in danger, but the assertion that the whole ladder lies above or below the real axis according to sin√(2γ) is not fully proven as written.\n\nSecond, the headline \\\"spectral singularity at any wavenumber\\\" is supported only by numerics for k ∈ (0,10]. The algorithm around (5.11) is explicit and the numerics plausible, but the abstract and conclusion state it as a theorem. Third, the monotone decrease of the PT-breaking threshold in Figure 5(a) comes from numerical continuation and is not proven. The forbidden gap in Lemma 11 is structurally sound, but the constant β* ≈ 4.808 is computed numerically, so that part is also partially computational.\n\nThe paper is honest about one limitation: it explicitly flags the omitted \\\"delicate analysis\\\" for sin√(2γ)=0, which is good practice, but it means that case is also open.\n\nThis is a paper for spectral theorists working on non-Hermitian operators and for the PT-symmetric photonics community. I would send it to a serious referee. It deserves conditional acceptance: the core existence and asymptotic results are worth publishing, but the authors should either prove the uniform remainder or explicitly soften the ladder-sign claim, and they should clearly label the \\\"any wavenumber\\\" and threshold-monotonicity results as numerical observations.","headline":"Solid PT-symmetric waveguide paper with a real new ladder effect, but the sign claim for the full ladder and two headline applications are not as proven as the abstract suggests.","tokens_in":28465,"tokens_out":6254,"would_cite":true,"duration_ms":60458,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81Q12","35P25","34L40"],"pacs":[],"model":"deepseek-v4-flash","headline":"Spacing the gain and loss elements of a PT-symmetric waveguide creates a ladder of resonances that can be converted to eigenvalues by tuning the gain-and-loss amplitude.","keywords":["PT symmetry","spectral singularities","resonances","eigenvalues","gain-and-loss distance","Wannier-Stark ladder","Schrödinger operator","laser-absorber modes"],"falsifier":"Numerically solve equation (2.9) at high precision for a fixed $\\gamma$ with $\\sin\\sqrt{2\\gamma}<0$, for several large values of $\\ell$, and for every integer $n$ satisfying (3.9), concentrating on the outermost values $n\\approx 2\\Theta(\\gamma)\\ell/\\pi$; if any of these roots has positive imaginary part, the claim that the whole ladder lies in the resonance half-plane fails.","tokens_in":27415,"feed_emoji":"⚛️","tokens_out":11180,"duration_ms":100077,"temperature":0.7,"pith_summary":"This paper studies a one-dimensional parity-time-symmetric (PT-symmetric) waveguide: two identical elements, one amplifying and one absorbing, separated by a variable distance. It tries to establish that the distance between gain and loss is a spectral switch, not just a geometric parameter. The central claim is that at large separation the system develops a ladder of resonances or of complex eigenvalues accumulating on a real segment of the spectrum, with the number of ladder members growing linearly with separation. Whether a given rung is a resonance or an eigenvalue is controlled by the sign of $\\sin\\sqrt{2\\gamma}$, where $\\gamma$ is the gain-and-loss amplitude, so a continuous sweep of $\\gamma$ converts a resonance ladder into an eigenvalue ladder. The paper also establishes that increasing separation lowers the PT-breaking laser-antilaser threshold and permits spectral singularities at any prescribed wavenumber.","feed_headline":"A wider gap gives PT-symmetric waveguides a resonance ladder","feed_subtitle":"At large separation the ladder's members crowd near the real axis, and tuning gain flips them between resonances and eigenvalues.","key_machinery":"The load-bearing object is the entire function $F(k,\\ell,\\gamma)$ of equation (2.9). It encodes the continuity conditions for a solution across the two elements: the product $F_-F_+$ describes the two isolated elements, while $e^{-4ik\\ell}F_0$ is the coupling through the gap, so the distance $\\ell$ enters only through the phase $e^{-4ik\\ell}$, making the spectral picture periodic in $\\ell$. Lemma 4's expansion (3.10), with roots located near $\\pi n/(2\\ell)$, converts this equation into the ladder statement, and the change of variables (3.17), reducing real roots to equation (5.2), makes the spectral-singularity analysis tractable. The explicit constant $\\Theta(\\gamma)$ in (3.9) sets the size of the accumulation segment.","core_discovery":"Working with a Schrödinger operator $-\\psi''+V(x)\\psi=k^2\\psi$ whose potential is a rectangular gain element at $x\\in(-\\ell-1,-\\ell)$ and its mirror-symmetric loss element, the authors reduce the spectral problem to a single equation $F(k,\\ell,\\gamma)=F_-(k,\\gamma)F_+(k,\\gamma)-e^{-4ik\\ell}F_0(k,\\gamma)=0$. Its roots $k$ describe all resonances, spectral singularities, and eigenvalues. For large $\\ell$, the roots near the real axis form ladders: Lemma 4 places exactly one simple root $k_n(\\ell,\\gamma)$ in each disk around $\\pi n/(2\\ell)$, and the four-term expansion (3.10) shows that the imaginary part is governed at order $\\ell^{-3}$ by $\\sin\\sqrt{2\\gamma}$. Hence the ladder consists of resonances when $\\sin\\sqrt{2\\gamma}<0$ and of eigenvalues when $\\sin\\sqrt{2\\gamma}>0$, and both types accumulate on the segment $[-\\Theta(\\gamma),\\Theta(\\gamma)]$ with $O(\\ell)$ members. In the same model, real roots of $F$, the spectral singularities, are analyzed through an auxiliary equation that allows one to engineer a singularity at any chosen real wavenumber and to make two distinct self-dual singularities coexist.","pith_inferences":["A testable extension the paper does not run: in a tunable photonic waveguide with fixed large $\\ell$, sweeping $\\gamma$ across a zero of $\\sin\\sqrt{2\\gamma}$ should flip the imaginary parts of the ladder from one side of the real axis to the other, converting resonances into eigenvalues.","The uniformity gap in the remainder suggests a targeted numerical check: track the outermost rung, $n\\approx 2\\Theta(\\gamma)\\ell/\\pi$, as $\\ell$ grows; its sign behaviour is where a failure of the ladder picture would first show up.","Because the distance enters the governing equation only through the phase $e^{-4ik\\ell}$, the same proof strategy should produce analogous ladders for other compactly supported gain-loss profiles $W(x)$, not just rectangular steps; the boundedness estimates in Lemma 4 are the only profile-dependent input."],"forward_implications":["At fixed gain-and-loss amplitude, increasing $\\ell$ creates more and more resonances or eigenvalues, with neighbouring rungs separated by distances of order $\\ell^{-1}$ and imaginary parts tending to zero.","A continuous change of $\\gamma$ across values where $\\sin\\sqrt{2\\gamma}$ changes sign transforms a ladder of resonances into a ladder of eigenvalues; at $\\sin\\sqrt{2\\gamma}=0$ the ladder consists of nearly spectral singularities with imaginary parts of order $\\ell^{-5}$.","Choosing the separation appropriately yields a spectral singularity at any prescribed real wavenumber, and two such singularities can be made to occur simultaneously at different wavenumbers.","The laser-antilaser (coherent-perfect-absorption) threshold decreases as the gain-to-loss distance grows, so spaced elements reach the lasing-absorbing regime at a smaller gain amplitude than adjacent elements.","There is a forbidden band of gain amplitudes, $\\pi^2/2<\\gamma<\\gamma_*\\approx 11.561$, in which no spectral singularities exist even though ladders are still present."],"supporting_citations":[{"why":"defines spectral singularities as zero-width resonances and supplies the laser/antilaser interpretation used throughout.","marker":"[1]"},{"why":"gives the norm-resolvent approximation for operators with distant perturbations, used to identify the $\\ell\\to\\infty$ limiting operator.","marker":"[21]"},{"why":"provides the general asymptotic-expansion scheme for distant localized perturbations that frames the ladder analysis.","marker":"[24]"},{"why":"is the Wannier-Stark resonance reference against which the large-distance ladder is compared.","marker":"[14]"},{"why":"establishes the phase transition through splitting of a self-dual spectral singularity, the mechanism used for the PT-breaking threshold.","marker":"[45]"},{"why":"states the conjecture of at most one spectral singularity for a parametrically fixed PT-symmetric potential that the two-singularity result contradicts.","marker":"[46]"}],"fun_headline_variants":["Gain-loss gap builds resonance ladders in PT waveguides","Separation spawns resonance ladders in PT-symmetric waveguides","Wide spacing flips PT waveguide resonances into eigenvalues","Gain-loss distance tunes resonance-to-eigenvalue ladders","Larger gap creates new resonances and eigenvalue ladders"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The ladder's sign rule rests on treating the remainder in expansion (3.10) as $O(\\ell^{-4})$ uniformly for every $n$ covered by condition (3.9), including $n$ proportional to $\\ell$; the paper proves the expansion pointwise in $n$ and does not show uniformity, so the outermost rungs of the ladder are the fragile ones.","fun_headline_variants_meta":{"raw":{"variants":["Gain-loss gap builds resonance ladders in PT waveguides","Separation spawns resonance ladders in PT-symmetric waveguides","Wide spacing flips PT waveguide resonances into eigenvalues","Gain-loss distance tunes resonance-to-eigenvalue ladders","Larger gap creates new resonances and eigenvalue ladders"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000537,"raw_usage":{"total_tokens":2605,"prompt_tokens":996,"completion_tokens":1609,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":612,"completion_tokens_details":{"reasoning_tokens":1524}},"tokens_in":612,"tokens_out":1609,"duration_ms":12221,"temperature":1.0,"reasoning_tokens":1524,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:47:15.836927+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Numerically solve equation (2.9) at high precision for a fixed $\\gamma$ with $\\sin\\sqrt{2\\gamma}<0$, for several large values of $\\ell$, and for every integer $n$ satisfying (3.9), concentrating on the outermost values $n\\approx 2\\Theta(\\gamma)\\ell/\\pi$; if any of these roots has positive imaginary part, the claim that the whole ladder lies in the resonance half-plane fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"defines spectral singularities as zero-width resonances and supplies the laser/antilaser interpretation used throughout."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the norm-resolvent approximation for operators with distant perturbations, used to identify the $\\ell\\to\\infty$ limiting operator."},{"cited_title":"Petersburg Math","cited_arxiv_id":null,"evidence_quote":"provides the general asymptotic-expansion scheme for distant localized perturbations that frames the ladder analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"is the Wannier-Stark resonance reference against which the large-distance ladder is compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"establishes the phase transition through splitting of a self-dual spectral singularity, the mechanism used for the PT-breaking threshold."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"states the conjecture of at most one spectral singularity for a parametrically fixed PT-symmetric potential that the two-singularity result contradicts."}],"review_version":1}