{"id":"494cb6d7-eab5-401b-8316-695e0b151673","arxiv_id":"2411.14776","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The non-Hermitian Kitaev chain is solved in the thermodynamic limit, giving the eigenvalue curves, the exact zero-mode condition, and the skin-effect criteria.","lead":"This paper gives exact formulas for the energy levels of a non-Hermitian version of the Kitaev chain, a model of a one-dimensional superconductor with gain and loss. It also maps out exactly when the chain hosts a zero-energy edge mode and when eigenstates pile up at the boundary (the skin effect).","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zero-mode criterion Eq. (29) depends on a branch choice in Sec. IV.B that the text admits may swap the x±,± values; the derivation should pin down the branch instead of deferring it.","rationale":"The reader flagged exactly this passage: the contradictory sentence in Sec. IV.B about the sign of 1/√(−4D2) and its claimed effect on the zero-mode region. I agree that this is the load-bearing weakness, because the zero-mode region is one of the three headline results, it is stated as a complete characterization in the abstract and in Sec. VII, and the derivation as written does not fix the branch. I considered the Vieta-equation characterization (Sec. V) as an alternative concern: it is supported by numerical plots and by the ordering constraint 1/|s| ≤ |κ|^2 ≤ |s|, and the derivation from |x2|=|x3| plus the resultant analysis is plausible; although a fully rigorous proof that the non-generic cases (double roots of a, t2=−t1, d1d2=0) are handled correctly is not written out, the reader's verdict already treats this as acceptable modulo revision. The skin-effect section is explicitly labeled as giving only sufficient conditions for complex parameters and the abstract's phrasing is softened there; that is a discrepancy between abstract and body but less damaging than the zero-mode branch ambiguity. The Bistritz analysis has no formal proof that all singular cases are exhausted, but the authors acknowledge this and the claim is appropriately hedged. Therefore the single most load-bearing concern remains the zero-mode branch ambiguity, and a concrete numerical scan of Eq. (29) against direct diagonalization, including a check of the two branch assignments, would settle whether the final formula is accidentally correct despite the sloppy derivation or genuinely incomplete. The verdict should remain CONDITIONAL: if the scan confirms Eq. (29), a revision that fixes the branch and removes the contradictory sentence would suffice for acceptance; if not, the zero-mode characterization would need substantive revision. I am not recommending REJECT because the rest of the paper's core method appears sound and independently checkable, and the concern is localized and testable.","tokens_in":24815,"tokens_out":2373,"duration_ms":20827,"concrete_test":"Perform a dense numerical scan over complex parameter space (e.g., random m,t1,t2,d1,d2 with |Re|,|Im| in [0,3], excluding only exact degeneracies), computing the two smallest |λ| of the open chain for L=200 with high-precision arithmetic. Mark a parameter point as having a zero mode if the smallest |λ| decays to zero with L (compare L=100,200,400). Then compare this numerical boundary against both (i) Eq. (29) as written, and (ii) Eq. (29) with the two arccos arguments exchanged and signs flipped to account for the alternative branch assignment, and also against (iii) the four direct conditions |x_{σ,σ'}|<1 using a fixed convention for √(4d1d2+m^2−4t1t2) and √(4d1d2+(t1−t2)^2). A discrepancy in any sector shows the derivation needs a branch-fixing lemma rather than a caveat. The same test should be run at d1d2=0 where Eq.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The abstract and Sec. VII claim that Eq. (54) [= Eq. (29)] 'fully' characterizes the zero-mode region for arbitrary complex parameters. The derivation in Sec. IV.B introduces the factor 1/√(−4D2) and states that this 'might introduce an additional sign for the square root terms', 'does influence the range of parameters for which there is a zero mode', and that the x values 'might be swapped' (text following Eq. (27)). These three statements are mutually inconsistent with a branch-independent final criterion. If the factor changes which of the four x±,± values are the small ones, then the two candidate conditions (|x−,−|,|x+,−| < 1) or (|x−,+|,|x+,+| < 1) can select different parameter regions, so Eq. (29), which is derived from only one assignment, is not established as the complete condition. The subsequent sentence 'the values of the various x±,± might be swapped' is presented as a harmless caveat, but the swap is exactly what determines which inequality must hold; a swapped assignment would replace |Im arccos(y1)| < |Im arccos(y2)| by the same inequality, so if the authors intended merely a relabeling the region would be unchanged, yet the text explicitly says the sign choice 'does influence the range of parameters'. The body does not resolve this contradiction: the finite-size checks in Sec. IV.B only illustrate d1d2→0 regularization and do not test Eq. (29) against a direct numerical zero-mode search in the complex parameter space where the ambiguity is live. Thus the central zero-mode claim is not yet backed by a branch-fixed derivation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the non-Hermitian Kitaev chain with arbitrary complex parameters m, t1, t2, d1, d2. The main results are: (i) a characterization of the eigenvalue curves in the infinite-size limit in terms of three Vieta equations (Eq. (37)) together with an ordering constraint 1/|s| ≤ |κ|^2 ≤ |s|; (ii) a criterion for the absence of skin effect, claimed to be complete for real parameters and conjectured sufficient for complex parameters; and (iii) a condition (Eq. (29)) that allegedly fully determines the parameter region in which a zero mode exists. The method combines a generalized Bloch ansatz, boundary determinant analysis, Vieta relations, and the Bistritz algorithm. The paper includes finite-size numerical checks (Figs. 1, 6, 7) supporting the Vieta-equation characterization.","tokens_in":25196,"tokens_out":7972,"duration_ms":69878,"significance":"If all three claims hold, the paper provides a valuable exact tool for non-Hermitian lattice models: a numerically stable characterization of the thermodynamic-limit spectrum, a zero-mode phase diagram, and a criterion for skin-effect absence. The Vieta-equation method is potentially generalizable to other one-dimensional non-Hermitian systems with longer-range couplings. A notable strength is the explicit comparison with finite-size numerics (Figs. 1, 6, 7), which demonstrates the practical utility of the eigenvalue characterization. However, the zero-mode section contains an unresolved internal contradiction about branch choices, and the skin-effect results for complex parameters are explicitly stated to be only conjectural. These points directly affect the paper's advertised completeness.","major_comments":[{"comment":"The derivation of the zero-mode criterion Eq. (29) is internally inconsistent. The text states that putting 1/√(−4D2) under the square roots 'might introduce an additional sign' and that 'this does influence the range of parameters for which there is a zero mode', but immediately adds that 'the values of the various x±,± might be swapped'. If the four x values are only permuted, the existence condition (two of them having |x|<1) is invariant, so the sign choice would not change the zero-mode region; if instead the sign choice can change which pair of x values is small, then Eq. (29), derived from one particular assignment, is not established as the full condition. The paper never resolves this contradiction. Moreover, the step from the conditions |x−,−|<1 ∧ |x+,−|<1 or |x−,+|<1 ∧ |x+,+|<1 to the inequality |Im arccos(y1)| < |Im arccos(y2)| is stated without a derivation ('Analysing these conditions, one finds...'). The finite-size checks at the end of Sec. IV.B address only the d1d2→0 regularization and do not test Eq. (29) against a direct numerical zero-mode search in the complex parameter region where the branch ambiguity is live. Thus the abstract's claim to 'fully determine the region in parameter space for which the model has a zero mode' is not supported by the presented derivation.","section":"Sec. IV.B"},{"comment":"The characterization of eigenvalues via the Vieta equations is stated to hold for 'arbitrary complex parameters', but the derivation explicitly excludes 'ten special values of λ' (plus the cases t2=-t1 and d1d2=0) with the remark 'We do not consider these ten special values, because we are interested the generic eigenvalues of the model'. The text does not show that these special values are not eigenvalues of the open chain in the L→∞ limit. If they are eigenvalues, they are missing from the claimed complete characterization; if they are not, a proof is needed. This is a load-bearing gap because the abstract and Sec. V claim a complete characterization of the eigenvalue curves.","section":"Sec. V"},{"comment":"For complex parameters, the skin-effect characterization is only conjectural. The text states: 'although we believe we determined all cases for which there are at least two roots on the unit circle, we do not have a proof for this in the general case with complex parameters' and later 'we believe that the conditions provided, exhaust all these cases.' The abstract, however, claims 'we characterise under which conditions the skin effect is absent' without specifying that this is only proven for real parameters. The paper should either limit the claim to the proven real-parameter case or provide a proof of necessity for the complex case; otherwise the abstract overstates the result.","section":"Sec. VI.B"}],"minor_comments":[{"comment":"In Table I, the condition 't2 = -t2' should read 't2 = -t1'.","section":"Table I"},{"comment":"The phrase 'Therefor' should be 'Therefore'.","section":"Sec. VI"},{"comment":"In Eq. (31), the parametrization x1 = s/κ, x2 = κ eiα, x3 = κ e−iα, x4 = 1/(sκ) would be clearer if it stated explicitly that α is real with 0 ≤ α < 2π, while κ and s are complex.","section":"Sec. V"},{"comment":"The definitions of y1 and y2 in Eq. (27) are missing punctuation in the typeset display, making them ambiguous. Please add a comma or semicolon between the two definitions.","section":"Sec. IV.B"}],"recommendation":"major_revision","confidential_remarks":"The paper presents an interesting and potentially useful method for non-Hermitian chains, and the Vieta-equation eigenvalue curves are convincingly validated numerically. However, the zero-mode section contains a genuine internal contradiction about the sign/branch choice that directly undermines the claim of a full characterization. The exclusion of 'ten special values' in Sec. V is also not justified. These issues are fixable, but they require substantive revision rather than mere copyediting. I would also encourage the authors to soften the abstract's skin-effect claim until the complex-parameter necessity is proven or clearly qualified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear X,\n\nThis paper gives the most complete analytic characterization I've seen of the thermodynamic-limit spectrum of the non-Hermitian Kitaev chain with arbitrary complex parameters. The main tool—requiring two of the four bulk roots x_i to have equal modulus (|x2|=|x3|)—is the generalized Brillouin zone condition from non-Bloch band theory, so the method isn't genuinely new, but the authors apply it cleanly and derive the three Vieta equations (37) plus the ordering constraint 1/|s| ≤ |κ|^2 ≤ |s|. The agreement with finite-size numerics (Figs. 1, 6, 7) is convincing, and the explanation of the branched eigenvalue curves in terms of which branch satisfies the ordering is a nice insight. The demonstration that machine-precision diagonalization becomes unstable for L=400 is a useful practical point.\n\nThe zero-mode condition (29)/(54) is the most interesting new result, but it's also the softest. The derivation in Sec. IV.B is compressed: 'Analysing these conditions, one finds' skips the key step, and the text around Eq. (27) contains a genuinely confusing statement. The authors say the factor 1/√(-4D2) 'might introduce an additional sign' and that this 'does influence the range of parameters for which there is a zero mode,' then immediately say the x±,± values 'might be swapped.' If the sign choice changes which pair of x's must be small, then the condition could be different, and Eq. (29) is not obviously the full answer. If it's just a simultaneous sign flip of y1 and y2, the inequality is invariant (as |Im arccos(-z)| = |Im arccos(z)|), but the text doesn't say that. Either way, the derivation needs to be fixed, and a direct numerical test of Eq. (29) across the complex parameter space would settle it. The finite-size check they do only probes the d1d2→0 regularization, which doesn't address the branch ambiguity.\n\nThe skin-effect section is fine for real parameters (m=0, t1=±t2, and the d1d2<0 case give a full set), but for complex parameters the abstract says 'characterise' while the text only gives sufficient conditions and 'believe' they are necessary. That's an overstatement. The statement in Sec. VI.B that T1(x)≡0 'seems' to exhaust the cases is not a proof.\n\nOverall, this is a worth-reading paper for anyone working on non-Hermitian topological models. The spectral characterization is probably right and will be used. The zero-mode result is likely right too, but it's not established as written. I'd send it to peer review, with the expectation of a revision that either pins down the branch choice or adds a numerical sweep supporting Eq. (29).","headline":"A solid analytic treatment of the non-Hermitian Kitaev chain spectrum, but the zero-mode criterion (Eq. 29) rests on a skipped derivation with a contradictory sign remark, and the abstract overstates the skin-effect results for complex parameters.","tokens_in":25665,"tokens_out":4476,"would_cite":true,"duration_ms":39958,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81Q12","82B23"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that the open non-Hermitian Kitaev chain's infinite-size spectrum is exactly the solutions of three Vieta equations with an ordering constraint, and that the zero-mode region is fixed by a single inverse-cosine inequality.","keywords":["non-Hermitian Kitaev chain","skin effect","zero mode","thermodynamic limit","Vieta equations","Bistritz algorithm","complex eigenvalues","Majorana zero modes"],"falsifier":"Take parameter sets near the boundary of the inequality (29), for instance with small nonzero $d_1, d_2$ and $m, t_1, t_2$ chosen so the two sides of the inequality are nearly equal, and compute the smallest eigenvalue modulus for $L = 100, 200, 400, 800$ using high-precision arithmetic; check that it decays to zero exactly on the predicted side of the boundary. The criterion is falsified if the boundary shifts when the square-root branch convention in Eq. (27) is changed, or if a zero-energy eigenvalue appears on the side predicted to have none.","tokens_in":24574,"feed_emoji":"⚛️","tokens_out":9465,"duration_ms":82873,"temperature":0.7,"pith_summary":"The paper studies the non-Hermitian Kitaev chain, a one-dimensional fermion model with complex hopping and pairing amplitudes, and aims to give a complete, numerically stable description of its open-chain spectrum in the infinite-size limit. It claims that every eigenvalue curve can be obtained from three Vieta equations for two complex parameters $\\kappa$ and $s$ and a real angle $\\alpha$, subject to the ordering constraint $\\frac{1}{|s|} \\leq |\\kappa|^2 \\leq |s|$, so the spectrum can be computed without the ill-conditioned diagonalisation of large non-Hermitian matrices. It also claims to fully determine the parameter region where a zero mode exists, via one inequality involving the imaginary parts of two inverse-cosine expressions built from the model parameters. Using these results, it characterises when the skin effect is absent and which eigenvectors are delocalised. The payoff is that three delicate questions—eigenvalue curves, zero modes, and skin-effect localisation—are reduced to algebraic conditions that can be checked directly.","feed_headline":"Three Vieta equations give the full non-Hermitian Kitaev spectrum","feed_subtitle":"A single arccos inequality also fixes the zero-mode region, and the skin effect is classified exactly for real parameters.","key_machinery":"The central object is the fourth-order bulk equation for the plane-wave factor $x$, obtained by substituting the ansatz $\\psi = (x, a x, x^2, a x^2, \\ldots)$ into the Bogoliubov-de Gennes equations. Because the quartic is palindromic, its four roots satisfy $x_1 x_2 x_3 x_4 = 1$. In the thermodynamic limit the boundary determinant is dominated by the roots of largest and second-largest modulus, and a non-trivial solution forces $|x_2| = |x_3|$; writing the roots as in Eq. (31) and applying the three remaining Vieta relations yields the eigenvalue equations (37). The zero-mode analysis uses the fact that at $\\lambda = 0$ the four roots factor into two reciprocal pairs, and the boundary conditions select the pair with both members inside the unit circle; the arccos inequality (29) encodes exactly when that happens. The skin-effect analysis uses the Bistritz algorithm, a root-counting algorithm for polynomials relative to the unit circle, to locate parameter choices for which two roots lie on the unit circle—the signature that the corresponding eigenstates are delocalised.","core_discovery":"In the limit $L \\to \\infty$, the eigenvalues $\\lambda$ of the open chain with arbitrary complex $m, t_1, t_2, d_1, d_2$ are exactly the solutions of the three Vieta equations (37), where the four roots of the bulk quartic are written as $x_1 = s/\\kappa$, $x_2 = \\kappa e^{i\\alpha}$, $x_3 = \\kappa e^{-i\\alpha}$, $x_4 = 1/(s\\kappa)$, with $\\kappa$ and $s$ complex and $0 \\leq \\alpha < 2\\pi$, and only solutions obeying $|x_1| \\geq |x_2| = |x_3| \\geq |x_4|$ (equivalently $\\frac{1}{|s|} \\leq |\\kappa|^2 \\leq |s|$) are physical eigenvalues. The zero mode $\\lambda = 0$ exists if and only if $\\left|\\operatorname{Im}\\arccos\\left(\\frac{-m}{2\\sqrt{t_1t_2 - d_1d_2}}\\right)\\right| < \\left|\\operatorname{Im}\\arccos\\left(\\frac{t_1+t_2}{2\\sqrt{t_1t_2 - d_1d_2}}\\right)\\right|$. In the hermitian limit this reduces to the familiar Kitaev conditions, while for $d_1d_2 = 0$ a zero mode can appear only at infinite system size. For real parameters the paper fully classifies the absence of the skin effect: it is absent when $m = 0$, when $t_1 = \\pm t_2$, or for the purely imaginary eigenvalue branch when $d_1d_2 < 0$ with the radicand negative; for complex parameters it gives sufficient conditions that it argues are also necessary. A further result is that the branch points visible in the spectrum arise from switching among three ordering branches of the Vieta solutions, with the non-physical branches explaining the geometry of the physical curves.","pith_inferences":["The same Vieta-branch method is likely transferable to other one-dimensional non-Hermitian models whose bulk polynomial is palindromic, since only the root-product identity and the ordering constraint are essential to the construction.","The zero-mode inequality may provide a direct non-Hermitian analogue of the topological phase boundary that could be compared with winding-number or biorthogonal-polarization invariants for the same model.","For complex parameters, the paper's skin-effect classification is only proven sufficient; if the stated belief that it is also necessary is wrong, there should exist parameter regions with extended $k$-intervals where the Bistritz algorithm is singular at level 4 but the phase conditions (52) fail.","The $d_1d_2 = 0$ case, where a zero mode appears only in the infinite-size limit, invites a closer check of whether the $L \\to \\infty$ limit commutes with the vanishing-pairing limit for boundary-driven phenomena."],"forward_implications":["The eigenvalue curves for any parameter set can be generated by numerically solving three polynomial equations over $\\alpha \\in [0, 2\\pi)$, avoiding the instability of brute-force diagonalisation of large non-Hermitian matrices.","The zero-mode region of the non-Hermitian Kitaev chain is fixed by a single explicit inequality, so the topological phase boundary can be drawn without solving the full spectrum.","The zero mode itself is not captured by the Vieta equations; it sits at $\\lambda = 0$ and is governed by the separate inequality (29), so a complete spectral picture needs both results.","The skin effect is absent exactly when two of the four bulk roots lie on the unit circle; for real parameters the paper lists all such cases, including the counterintuitive case of real eigenvalues with skin effect.","Phase-rotating the parameters rigidly rotates the entire spectrum, giving a one-parameter family of parameter sets with identical eigenvalue structure up to an overall phase."],"supporting_citations":[{"why":"Defines the Kitaev chain model and its hermitian Majorana zero modes, the baseline that the non-hermitian generalisation extends.","marker":"[66]"},{"why":"Supplies the plane-wave ansatz for solving bulk recurrence relations in one-dimensional fermion chains.","marker":"[87]"},{"why":"Provides the exact finite-size solution method for the Kitaev chain with longer-range couplings, adapted here to the case $t_1 = t_2$ with complex parameters.","marker":"[88]"},{"why":"Identifies the $d_1d_2 = 0$ limit of the model with the Hatano-Nelson non-reciprocal hopping model and its localisation physics.","marker":"[90]"},{"why":"Supplies techniques for non-hermitian systems that interpolate between open and periodic boundary conditions, used for the eigenvalues when $d_1d_2 = 0$.","marker":"[65]"},{"why":"Defines the Bistritz algorithm used to count roots inside, on, and outside the unit circle, which is the core tool for characterising the skin effect.","marker":"[97]"},{"why":"Provides the non-Bloch band theory language that connects skin-effect-free eigenvectors to real momenta and skin-effect eigenvectors to complex momenta.","marker":"[60]"},{"why":"States the Vieta relations between polynomial coefficients and roots, which form the backbone of the three eigenvalue equations (37).","marker":"[93]"}],"fun_headline_variants":["Vieta triple yields exact non-Hermitian Kitaev spectrum","One arccos inequality fixes the zero-mode condition","Skin effect absence classified for real Kitaev parameters","Eigenvalue curves solved by Vieta branch ordering","New Vieta method cracks non-Hermitian Kitaev chain"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The zero-mode criterion Eq. (29) assumes that the complex square-root factor introduced in Eq. (27) can only swap the labels of the four $x$ solutions, not change the parameter region where a zero mode exists; the text itself remarks that the sign might introduce an additional sign and that it 'does influence' the range, so if that caveat is literal the criterion may not be complete.","fun_headline_variants_meta":{"raw":{"variants":["Vieta triple yields exact non-Hermitian Kitaev spectrum","One arccos inequality fixes the zero-mode condition","Skin effect absence classified for real Kitaev parameters","Eigenvalue curves solved by Vieta branch ordering","New Vieta method cracks non-Hermitian Kitaev chain"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00018,"raw_usage":{"total_tokens":1359,"prompt_tokens":1055,"completion_tokens":304,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":671,"completion_tokens_details":{"reasoning_tokens":225}},"tokens_in":671,"tokens_out":304,"duration_ms":3746,"temperature":1.0,"reasoning_tokens":225,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:54:46.542415+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take parameter sets near the boundary of the inequality (29), for instance with small nonzero $d_1, d_2$ and $m, t_1, t_2$ chosen so the two sides of the inequality are nearly equal, and compute the smallest eigenvalue modulus for $L = 100, 200, 400, 800$ using high-precision arithmetic; check that it decays to zero exactly on the predicted side of the boundary. The criterion is falsified if the boundary shifts when the square-root branch convention in Eq. (27) is changed, or if a zero-energy eigenvalue appears on the side predicted to have none.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the Kitaev chain model and its hermitian Majorana zero modes, the baseline that the non-hermitian generalisation extends."},{"cited_title":"Rahul, N","cited_arxiv_id":null,"evidence_quote":"Supplies the plane-wave ansatz for solving bulk recurrence relations in one-dimensional fermion chains."},{"cited_title":"Zhou, Entanglement Phase Transitions in Non-Hermitian Kitaev Chains , Entropy 26, 272 (2024)","cited_arxiv_id":null,"evidence_quote":"Provides the exact finite-size solution method for the Kitaev chain with longer-range couplings, adapted here to the case $t_1 = t_2$ with complex parameters."},{"cited_title":"Fukui, Y","cited_arxiv_id":null,"evidence_quote":"Identifies the $d_1d_2 = 0$ limit of the model with the Hatano-Nelson non-reciprocal hopping model and its localisation physics."},{"cited_title":"Zirnstein, G","cited_arxiv_id":null,"evidence_quote":"Supplies techniques for non-hermitian systems that interpolate between open and periodic boundary conditions, used for the eigenvalues when $d_1d_2 = 0$."},{"cited_title":"Hatano, D","cited_arxiv_id":null,"evidence_quote":"Defines the Bistritz algorithm used to count roots inside, on, and outside the unit circle, which is the core tool for characterising the skin effect."},{"cited_title":"Herviou, J","cited_arxiv_id":null,"evidence_quote":"Provides the non-Bloch band theory language that connects skin-effect-free eigenvectors to real momenta and skin-effect eigenvectors to complex momenta."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the Vieta relations between polynomial coefficients and roots, which form the backbone of the three eigenvalue equations (37)."}],"review_version":1}