{"id":"fdb29d99-203c-430d-a84d-5245d7d75d55","arxiv_id":"2505.20791","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In a non-Hermitian SSH chain with non-reciprocal hopping, staggered gain/loss tunes defect-state localization between the defect site, the chain edges, or suppression.","lead":"This paper shows that adding staggered gain and loss to a non-Hermitian chain can switch where a defect state sits: at the defect, at the edges, or not at all. It proposes a topolectrical circuit to realize this switching, which could matter for reconfigurable signal processing and sensing.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (40) grows as |β_>β_<|^{-(n+1)} with |β_>β_<|<1 for δ<0, so the paper's own formulas predict right-edge dominance for large n, not defect localization; the three-regime claim needs an n-controlled re-derivation.","rationale":"The reader's weakest assumption correctly identifies the large-n approximation as the delicate point. My stress-test sharpens this into a concrete algebraic inconsistency: for the chosen δ<0, the product |β_>β_<| is smaller than 1, so the right-edge amplitude in Eq. (40) grows exponentially with n at a rate −ln|β_>β_<| while the defect amplitude grows at a rate ln|β_>|; for large n the edge term wins. This means the claimed exponential dominance of the defect site is not a consequence of the displayed formulas, and the analytical criterion based on |β_<| and |β_>| crossing unity is not sufficient. The exact-diagonalization and LTSpice results in Figs. 4 and 6 do provide independent evidence for the qualitative three-regime picture at the sizes studied, so I am not moving the verdict to REJECT. However, the manuscript should either correct the derivation, prove a bound on the dropped terms, or explicitly state the n range over which the defect-localized regime survives. The reader's CONDITIONAL verdict remains the appropriate outcome; my concern strengthens the reasons for the conditions rather than changing the verdict.","tokens_in":19810,"tokens_out":18029,"duration_ms":193452,"concrete_test":"For the parameters of Fig. 4 (t1=1.2, t2=4, δ=−1.5), compute the unnormalized zero-energy eigenvector from the exact recurrence and the three defect equations for n=10, 20, and 40, and compare |ψ(x=2n)|, |ψ(x=1)|, and |ψ(x=4n−1)| with Eqs. (38)-(40), including the terms β_<^{n-1} and β_>^{-(n-1)} that were dropped. If the ratio |ψ(2n)|/|ψ(4n−1)| at γ=1.5 is not exponentially larger for large n—or if the defect-localized interval in γ does not track the β-crossing conditions—then the three-regime claim is a finite-size artifact and the analytical derivation must be revised before the central claim is accepted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The finite-γ defect-state derivation in the section 'EFFECT OF GAIN/LOSS ON THE DEFECT STATES' constructs the zero-energy wavefunctions from the bulk β roots. The central prediction—that in the PT and APT regimes the zero-energy weight is exponentially concentrated at the defect—is justified by neglecting β_<^{n-1} and β_>^{-(n-1)} in Eqs. (36)-(37) and by asserting that the defect amplitude (39) is exponentially larger than the edge amplitudes (38) and (40). This assertion does not follow for the paper's own parameters. From the zero-energy quadratic in β, the product of the two roots is β_>β_< = (t2+δ)/(t2−δ), which for t2=4, δ=−1.5 has magnitude 0.4545 < 1. Equation (40) for the rightmost node contains 1/(β_>β_<)^{n+1} = (5.5/2.5)^{n+1}, which grows exponentially with n rather than being suppressed. Thus the statement after Eq. (40) that the defect wavefunction becomes exponentially larger than the wavefunctions at the ends is inconsistent with the displayed formulas unless the prefactor α_</α_> cancels this growth; for γ=1.5 this prefactor is only about 0.034, which does not cancel the exponential growth at large n. Consequently, the localization condition cannot be read off from |β_<| and |β_>| alone: the α-dependent defect-matching coefficients determine the exponential hierarchy, and the claimed three-regime classification and critical values γc1...γc4 are not established analytically. The numerical and LTSpice results may still support the qualitative picture, but the analytical derivation must be repaired or its regime of n validity stated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a finite non-Hermitian SSH chain with non-reciprocal inter-cell hopping and staggered gain/loss, including a central defect site. The authors show that in the absence of gain/loss the non-Hermitian skin effect drags both bulk and edge states to one edge, and that staggered gain/loss restores the two edge states to opposite edges. For the defect state, they identify three regimes controlled by the gain/loss strength γ: localization at the defect site in the PT and anti-PT regimes, localization at the system edges in the broken-PT regime, and a claimed complete delocalization/suppression in the broken-PT regime. The analytical argument uses the non-Bloch factors β< and β> obtained from the zero-energy bulk condition, with the defect wavefunction constructed from the corresponding eigenvectors. The paper closes with an LTSpice-based topolectrical circuit realization and a comparison of voltage and admittance spectra with the numerical model.","tokens_in":20124,"tokens_out":12874,"duration_ms":126727,"significance":"If the central claim holds, the work provides a concrete route to reconfigurable defect-state engineering in non-Hermitian systems, extending the earlier defect activation/suppression picture of Stegmaier et al. to models with non-reciprocal hopping and the non-Hermitian skin effect. The manuscript is unusually transparent in presenting its analytical construction, and the numerical and LTSpice data are consistent with each other for the system sizes shown. The model has no fitted free parameters, and the circuit proposal is plausible. However, the analytical derivation of the three-regime classification contains a size-dependent asymptotic error that directly affects the paper's central claim, so the presented evidence is not yet sufficient as it stands.","major_comments":[{"comment":"The statement after Eq. (40) that, at large n, the defect amplitude (39) becomes exponentially larger than the right-end amplitude (40) is inconsistent with the displayed formulas for the paper's own parameters. The zero-energy quadratic gives β<β> = (t2+δ)/(t2−δ), so for t1=1.2, t2=4, δ=−1.5 one has |β<β>| = 2.5/5.5 ≈ 0.455 and |1/(β<β>)|^{n+1} ≈ (2.2)^{n+1}. Equation (40) therefore grows exponentially with n, while Eq. (39) grows as |β>|^n ≈ (1.69)^n for the same parameters. The prefactor α</α> ≈ 0.034 does not remove this growth; for large enough n the right-edge amplitude dominates. The claimed exponential hierarchy and the resulting three-regime classification cannot be read off from |β<| and |β>| alone; the matching coefficients α<, α>, and c>,R, together with the n-dependent factor in Eq. (40), determine the localization. The authors should re-derive the full matching conditions, identify the actual condition under which the defect amplitude dominates the right-edge amplitude (e.g. involving |β>|^2|β<| and the prefactors), and test numerically for larger n than n=10 whether the defect-localized regime persists.","section":"EFFECT OF GAIN/LOSS ON THE DEFECT STATES, Eqs. (38)-(40)"},{"comment":"The critical values γc1...γc4 are derived purely from the conditions |β<|=1, |β>|=1, and |β<|=|β>|, which are properties of the bulk generalized Brillouin zone. The localization of the defect zero mode is, however, governed by the defect-matching problem, including the boundary condition on both sides of the defect and the normalization of the full wavefunction. Since the matching coefficients in Eqs. (38)-(40) are n-dependent, the points where the defect amplitude crosses the edge amplitudes generally differ from the bulk β-crossings. Without an explicit comparison of the full amplitudes, the identification of γc1...γc4 as the boundaries of the three localization regimes is not established. The authors should either prove that the matching coefficients do not affect the exponential hierarchy in the regimes they claim, or replace the γc values by size-dependent crossover values computed from the exact wavefunctions.","section":"EFFECT OF GAIN/LOSS ON THE DEFECT STATES, Fig. 5d and text defining γc1...γc4"},{"comment":"The abstract states that in the three regimes the defect states 'localize at the defect site, shift to the system's edges, or become completely delocalized.' The main text and Fig. 4e instead show that in the broken-PT regime the defect state is absent or suppressed, not spread over the chain, and the conclusion similarly says 'suppress the localization of the defect state completely.' This is an inconsistency in the central claim as stated: 'completely delocalized' is not the same as 'absent.' The authors should clarify what is meant in the broken-PT regime, and adjust the abstract and conclusion accordingly.","section":"Abstract and Conclusion"}],"minor_comments":[{"comment":"The stated condition δ²+γ²−t1²−t2²<0 is violated for the APT case shown in Fig. 4, e.g. γ=5.5 with the stated parameters gives δ²+γ²−t1²−t2²≈15.06>0. Please specify how the β roots are obtained in that regime (e.g. analytic continuation or a different branch choice).","section":"Eq. (32)"},{"comment":"The text says the white line denoting high probability density shifts to the defect site when γ exceeds approximately 0.5×10⁻⁶, but the horizontal axis of Fig. 5b appears to be γ on a scale of 0 to 6. The quoted threshold seems inconsistent with the figure scale or the stated parameters; please clarify.","section":"Fig. 5b"},{"comment":"There are several typos and OCR-style artifacts in the equations and text, e.g. 'Schroedinger', the duplicated '2' in the denominator of Eq. (32), and the garbled rendering of Eq. (39). A careful proofreading pass would improve readability.","section":"General presentation"},{"comment":"The circuit validation is a simulation (LTSpice) of the same model, not an independent experimental measurement. The wording is mostly careful, but the phrase 'validate our theoretical predictions' in the conclusion could be read as claiming more than a consistency check. Please rephrase to emphasize that the circuit is a proposed experimental realization supported by circuit simulations.","section":"Circuit section"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is likely of interest to the non-Hermitian topological circuits community, and the numerical and LTSpice results are consistent for the system sizes shown. The main issue is the asymptotic claim after Eq. (40), which is demonstrably wrong for the paper's own parameters: the right-edge amplitude grows faster in n than the defect amplitude when the product |β<β>| is less than unity. This is a load-bearing error in the analytic derivation of the three-regime classification. The authors should either repair the analytic argument with full matching conditions or substantially qualify the claim to finite system sizes and show that the regime boundaries are size-dependent. The paper is not in a state that I can recommend for acceptance in its current form, but the underlying numerical model and the proposed circuit are promising enough that a major revision is appropriate. I also note a high degree of self-citation in the reference list; this is not a reason to reject, but the authors may wish to cite more broadly the non-Hermitian defect-state literature beyond their own group's work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You asked for my take on this paper. The new thing here is the three-regime picture: defect states sit at the defect in the PT and APT regimes, move to the edges in the broken-PT regime, with crossover values gamma_c1...gamma_c4. That extends Stegmaier and prior SSH defect work by adding non-reciprocal hopping, and the LTSpice simulations match the numerics well. For people in the non-Hermitian circuits subfield, the numerics and the circuit design are useful.\n\nThe soft spots are real, though. The abstract says \"completely delocalized\" for the broken-PT regime, but the actual profile is edge-localized, not delocalized. That is an overclaim.\n\nMore serious is the analytical derivation. The authors claim that at large n the defect amplitude grows as beta_>^n and therefore dominates the edges. But Eq. (40) for the rightmost node contains (beta_> beta_<)^{-(n+1)}, and for their parameters (t1=1.2, t2=4, delta=-1.5) the product has magnitude about 0.45. So that term also grows exponentially with n, with an exponent larger than that of beta_>^n. The small prefactor alpha_</alpha_> (about 0.03) can keep the right edge smaller at n=10, which is likely why the numerics look fine, but the asymptotic statement \"exponentially larger\" is not supported by their own formulas. The regime boundaries derived from |beta| crossing unity are not enough when the alpha-dependent prefactors control the exponential hierarchy. The large-N approximations in Eqs. (36)-(37) also lack quantified error bounds.\n\nNone of this kills the qualitative message: the numerics and LTSpice consistently show defect localization at the sizes studied. But the analytical justification needs repair, or at least an explicit statement of the n-range where it is valid. The LTSpice simulation is a simulation of the same model, so it is a consistency check rather than an independent experiment, but I do not see circularity in the derivation itself.\n\nWho gets value: experimentalists and theorists working on non-Hermitian topological circuits and PT-symmetric defect engineering. It deserves a serious referee, and I would send it to review, but with a request for major revision of the analytical part and a corrected abstract.","headline":"Worth refereeing for its numerical and circuit-simulation study of defect states under combined NHSE and PT symmetry, but the analytical localization argument has a large-n flaw that needs repair.","tokens_in":20745,"tokens_out":4939,"would_cite":false,"duration_ms":47726,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In a non-Hermitian SSH chain, staggered gain and loss alone can park a defect state at the defect site, push it to either edge, or erase it.","keywords":["non-Hermitian skin effect","parity-time symmetry","defect states","Su-Schrieffer-Heeger model","topolectrical circuits","gain and loss","topological edge states","non-Bloch band theory"],"falsifier":"Exact diagonalization of the finite defective chain at several sizes ($n=5,10,20$) can be checked against the predicted peak position of the zero-energy state: it should sit at the defect node for the PT and APT regimes, move to an edge when the $\\beta$-magnitude conditions fail, and show no defect-localized zero mode in the broken-PT window. A mismatch between the numerically obtained peak and the $\\beta$-based boundaries would falsify the analytical mechanism, as would a topolectrical-circuit measurement that shows voltage accumulation at the defect node in the broken-PT regime.","tokens_in":19581,"feed_emoji":"⚡","tokens_out":13158,"duration_ms":123480,"temperature":0.7,"pith_summary":"The paper sets out to show that in a finite non-Hermitian Su-Schrieffer-Heeger (SSH) chain with non-reciprocal inter-cell hopping and balanced gain and loss on the two sublattices, the strength $\\gamma$ of the gain/loss alone controls where a topological defect state lives. Three regimes are identified: in the parity-time (PT) symmetric and anti-PT (APT) symmetric regimes the zero-energy defect state sits at the defect node, while in the broken-PT regime it is no longer defect-localized and instead shifts toward the chain edges, where it is suppressed. The authors provide an analytic description based on two non-Bloch decay factors extracted from the bulk, including closed-form expressions for the four critical $\\gamma$ values at which the localization switches, and they validate the picture with a topolectrical-circuit proposal whose LTSpice voltage and admittance simulations reproduce the predicted signatures. A sympathetic reader would care because the result turns defect-state engineering into a single-knob control problem on a tabletop electrical platform.","feed_headline":"Staggered gain and loss repositions a topological defect state","feed_subtitle":"One knob moves a zero-energy state between the defect site, the chain edges, and nowhere.","key_machinery":"The load-bearing object is the pair of non-Bloch factors $\\beta_<$ and $\\beta_>$, the generalized Bloch roots (decay/growth factors) of the bulk characteristic equation for the zero-energy manifold, together with the sublattice-weight amplitudes $\\alpha_<$ and $\\alpha_>$. On either side of the defect, the zero-energy wavefunction is written as a combination of $\\beta_<^m$ and $\\beta_>^m$ components, and the condition for the state to be pinned at the defect is that the left factor decays rightward and the right factor decays leftward, i.e. $|\\beta_<|<1$ and $|\\beta_>|>1$. The critical values $\\gamma_{c1},\\gamma_{c2},\\gamma_{c3},\\gamma_{c4}$ mark precisely where $|\\beta_<|$ or $|\\beta_>|$ crosses unity or where the two factors become equal, so the whole three-regime classification follows from tracking these two complex-plane radii.","core_discovery":"On its own terms, the paper's central claim is that the zero-energy defect state in a finite non-Hermitian SSH chain is reconfigurable by staggered gain/loss: depending on $\\gamma$ it localizes at the defect site, shifts to an edge of the chain, or becomes completely delocalized. The control mechanism is the magnitude ordering of the two non-Bloch factors $\\beta_<$ and $\\beta_>$ obtained from the bulk characteristic equation; defect-localized wavefunctions require $|\\beta_<|<1$ and $|\\beta_>|>1$, and the four critical values $\\gamma_{c1} = \\sqrt{(|t_2|-|t_1|)^2-\\delta^2}$, $\\gamma_{c2} = \\sqrt{t_2^2-\\delta^2}-|t_1|$, $\\gamma_{c3} = \\sqrt{t_2^2-\\delta^2}+|t_1|$, and $\\gamma_{c4} = \\sqrt{(|t_1|+|t_2|)^2-\\delta^2}$ are exactly the points where these inequalities change. The paper further claims that introducing gain/loss restores the two topological edge states to opposite ends of the chain even while the bulk non-Hermitian skin effect persists, and that defect states survive only in the PT- and APT-symmetric regimes.","pith_inferences":["The mechanism is stated in terms of generic $\\beta$ factors, so the same three-regime classification should carry over to other one-dimensional non-Hermitian bipartite lattices with a domain wall and a sublattice-pinned zero mode, including photonic or mechanical arrays with balanced loss.","Because the defect amplitude grows roughly as $|\\beta_>|^n$, the switch near $\\gamma_{c1}$ and $\\gamma_{c4}$ should become sharper as the chain grows, so finite-size effects in the proposed 39-node circuit are a natural place to test the approximation.","The model leaves the defect node passive; adding gain or loss on the defect itself is a direct variation that would reveal how much of the zero-energy pinning depends on the defect remaining unpaired."],"forward_implications":["Tuning $\\gamma$ across the four critical values acts as a three-position switch for a topological zero mode: at the defect site, at an edge, or absent.","In the proposed topolectrical circuit, the defect node's voltage and zero-admittance accumulation should appear only in the PT and APT regimes, giving an electrical readout of the phase.","Because the bulk states remain NHSE-localized at one edge, the restored edge states at opposite ends can be read out separately from the bulk background.","The closed-form expressions for $\\gamma_{c1},\\ldots,\\gamma_{c4}$ mean the regime boundaries are predictable from the circuit capacitances and resistances before fabrication."],"supporting_citations":[{"why":"Sets the PT/APT/BPT defect-state activation scenario in non-Hermitian electrical circuits that this work extends by adding non-reciprocal hopping.","marker":"[68]"},{"why":"Provides the generalized Lieb-theorem result used to pin the defect states at zero energy.","marker":"[75]"},{"why":"Supplies the non-Hermitian edge-state description used to interpret how gain/loss moves the two edge states.","marker":"[72]"},{"why":"Shows how INIC-based non-reciprocal elements realize asymmetric couplings in topolectrical circuits, underpinning the proposed experiment.","marker":"[38]"},{"why":"Gives a second realization of INIC-based asymmetric coupling used in the circuit proposal.","marker":"[71]"}],"fun_headline_variants":["Gain and loss repositions a defect state to site, edge, or nowhere","Three regimes: defect state hops from site to edge to delocalized","Staggered gain-loss steers defect state among three localizations","One knob tunes a defect state: site, edge, or delocalize","Defect state reconfigurable by staggered gain and loss"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The classification assumes that the finite-chain defect wavefunction is faithfully described by the two bulk non-Bloch factors $\\beta_<$ and $\\beta_>$, with terms of order $\\beta_<^{n-1}$ and $\\beta_>^{-(n-1)}$ dropped; if the defect itself renormalizes these factors, the critical $\\gamma$ values and the three-regime picture would shift.","fun_headline_variants_meta":{"raw":{"variants":["Gain and loss repositions a defect state to site, edge, or nowhere","Three regimes: defect state hops from site to edge to delocalized","Staggered gain-loss steers defect state among three localizations","One knob tunes a defect state: site, edge, or delocalize","Defect state reconfigurable by staggered gain and loss"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001458,"raw_usage":{"total_tokens":5945,"prompt_tokens":1097,"completion_tokens":4848,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":713,"completion_tokens_details":{"reasoning_tokens":4753}},"tokens_in":713,"tokens_out":4848,"duration_ms":29652,"temperature":1.0,"reasoning_tokens":4753,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T13:46:29.013995+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exact diagonalization of the finite defective chain at several sizes ($n=5,10,20$) can be checked against the predicted peak position of the zero-energy state: it should sit at the defect node for the PT and APT regimes, move to an edge when the $\\beta$-magnitude conditions fail, and show no defect-localized zero mode in the broken-PT window. A mismatch between the numerically obtained peak and the $\\beta$-based boundaries would falsify the analytical mechanism, as would a topolectrical-circuit measurement that shows voltage accumulation at the defect node in the broken-PT regime.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Sets the PT/APT/BPT defect-state activation scenario in non-Hermitian electrical circuits that this work extends by adding non-reciprocal hopping."},{"cited_title":"Zurita, C","cited_arxiv_id":null,"evidence_quote":"Provides the generalized Lieb-theorem result used to pin the defect states at zero energy."},{"cited_title":"Rafi-Ul-Islam, Z","cited_arxiv_id":null,"evidence_quote":"Supplies the non-Hermitian edge-state description used to interpret how gain/loss moves the two edge states."},{"cited_title":"Sahin, Z","cited_arxiv_id":null,"evidence_quote":"Shows how INIC-based non-reciprocal elements realize asymmetric couplings in topolectrical circuits, underpinning the proposed experiment."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives a second realization of INIC-based asymmetric coupling used in the circuit proposal."}],"review_version":1}