{"id":"2f280f42-d0b6-494b-a4f4-7a8c9f299b63","arxiv_id":"2411.14328","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Strong loss-gain in a staggered non-Hermitian SSH lattice turns all insulators into non-trivial topological insulators, passing through a gapless semi-metallic phase.","lead":"This paper studies a one-dimensional chain where every other pair of sites has equal gain and loss added. It finds that strong gain-loss pushes the system into a topologically non-trivial insulating state in every parameter region, via an unusual gapless intermediate phase.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'non-trivial semi-metallic' label is inferred from a Zak phase that is not quantized at exceptional points, so the distinctive transition claim is not yet supported.","rationale":"The reader identified the semimetal Zak phase as the weakest assumption. I agree: the paper's headline phenomenon depends on labeling the gapless phase non-trivial, and Eq. (16) is not applicable across exceptional points. The rest of the argument—critical u values matching band closings, strong-u Zak phase, edge states—appears internally consistent. The concrete test with a complex-energy winding number or branch-cut-resolved Zak phase would settle this without requiring a full re-derivation. If the test yields ν = 0, the verdict should remain CONDITIONAL with the claim weakened; if ν ≠ 0, the claim is supported. No change to the reader's CONDITIONAL verdict is needed.","tokens_in":14815,"tokens_out":5074,"duration_ms":49028,"concrete_test":"For the gapless region (e.g., θ = π/4, u = 1.5), compute a well-defined invariant for non-Hermitian bands: the winding number ν = (1/2π) ∮_BZ d/dk arg det[H(k) − E_ref] dk for a reference point E_ref in a complex-energy gap, or the sum of Zak phases on intervals between consecutive exceptional points with explicit branch cuts. If ν = 0, or if the interval-summed Zak phase depends on the placement of branch cuts, the semimetal is not topologically non-trivial, and the central claim should be revised to describe two non-trivial insulators separated by an unclassified gapless phase.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the gapless semi-metallic phase is 'non-trivial' relies on the biorthogonal Zak phase of Eq. (16) evaluated across the whole BZ in Sec. 3.2 (Fig. 4). For uc1 < u < uc2, the bands touch at exceptional points (Eqs. (10), (12)), where the eigenstates coalesce and the eigenvector frame is not continuous. A Zak phase computed through such a region is gauge-dependent and not a quantized invariant; the text concedes that it 'fluctuates around 2π'. Therefore the abstract's assertion that the transition is mediated by a 'non-trivial semi-metallic phase' is not established by the presented calculation. The sequence insulator-gapless-insulator is verified by the band structure and is not in question; what is in question is the topological label attached to the gapless interval. If the semimetal cannot be assigned a well-defined non-trivial invariant, the 'unusual transition' reduces to a gap-closing-reopening process between two non-trivial insulators, which is less distinctive than claimed. A secondary issue: the abstract's claim of 'only complex eigenspectra for all u ≠ 0' is contradicted by real-spectrum intervals visible in Fig. 2(c), but this does not affect the phase-transition argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies a one-dimensional non-Hermitian Su-Schrieffer-Heeger (SSH) model with four sites per unit cell, in which two sites carry balanced loss and gain while the other two are Hermitian. The authors derive the four-band Bloch spectrum, identify two exceptional-point mechanisms, and obtain closed-form critical values u_c1, u_m, and u_c2 that separate insulating and semi-metallic intervals. Using a biorthonormal Zak phase, they conclude that for sufficiently strong loss-gain strength the system always enters a non-trivial insulating phase, even when the Hermitian limit is trivial, and that a non-trivial Hermitian insulator passes through a gapless semi-metallic phase before returning to a non-trivial insulating phase. They also analyze open-boundary spectra, edge modes, inverse participation ratios, and robustness to disorder.","tokens_in":15028,"tokens_out":11570,"duration_ms":106143,"significance":"If the phase diagram is correct, the model provides a simple analytically tractable example in which non-Hermiticity restores non-trivial topology in the strong-coupling limit and in which a topological gap closing is not accompanied by a change of the gapped-phase invariant. The closed-form expressions for the critical lines and the independent numerical confirmation are strengths, as is the OBC/IPR analysis that distinguishes the two non-trivial insulating regimes. However, the paper's most distinctive claim, that the gapless intermediate phase is itself 'non-trivial', is not supported by the presented invariant because the Zak phase is not quantized in the presence of exceptional points; in addition, the abstract contains a spectral statement contradicted by the authors' own band-structure results.","major_comments":[{"comment":"The claim that the system 'supports only complex eigenspectra for all values of u ≠ 0' is contradicted by the authors' own band-structure analysis. In Sec. 3.1, for θ = π/4 and u = 1.25 (which lies between u_c1 and u_m), the text states that 'completely real energies exist only between EP 1 and EP 2', and Fig. 2(c) correspondingly shows extended intervals of real spectrum. The spectrum is therefore partially complex, not exclusively complex. This statement appears in the abstract and in the Fig. 5 caption and must be corrected, since it is presented as a headline result.","section":"Abstract; Sec. 3.1 after Eq. (12); caption of Fig. 5"},{"comment":"The topological label 'non-trivial' attached to the gapless semi-metallic phase is not established. In the interval u_c1 < u < u_c2 the bands touch at exceptional points (Eqs. (10) and (12)), where eigenstates coalesce and the biorthonormal frame used in Eq. (16) is not a continuous frame over the Brillouin zone. A Zak phase computed through such a region is gauge-dependent and is not a quantized invariant. The text itself concedes that Ω 'fluctuates around 2π' in this region. Fluctuating values cannot certify a non-trivial invariant. Since the abstract's central claim is that the transition is mediated by a 'non-trivial semi-metallic phase', this needs either a well-defined non-Hermitian invariant valid at exceptional points (for example, a winding of the complex-energy spectrum around the EPs) or revised wording that limits the claim to a gapless intermediate region whose topology is not classified by the Zak phase. The insulator-gapless-insulator sequence and the non-trivial labels of the two gapped phases are not in question.","section":"Sec. 3.2, Eq. (16), Fig. 4; Sec. 5"}],"minor_comments":[{"comment":"The caption says the OBC/PBC comparison is for θ = π/4, but the surrounding text and the displayed data correspond to θ = 3π/4.","section":"Caption of Fig. 8"},{"comment":"The bracket structure in H_hop is inconsistent: the square bracket opens before the two bra-ket terms but closes after 'h.c.', leaving the parentheses unbalanced.","section":"Eq. (3)"},{"comment":"The text refers to the 'Kiteav chain'; this should be 'Kitaev chain'.","section":"Sec. 1"},{"comment":"The manuscript should specify which bands are included in the sum in Eq. (16) and state the gauge and branch choices used for the multi-valued square roots in Eq. (9), so that the numerical Zak-phase values can be independently reproduced.","section":"Sec. 3.2"},{"comment":"The claim that edge states in the gapless region are robust is not directly illustrated: Fig. 10 shows disorder robustness only for u = 0.5 and u = 3.5, which are gapped-phase values, not for a representative point inside the semi-metallic interval.","section":"Sec. 4, Fig. 10"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is honest in reporting that the Zak phase fluctuates around 2π in the gapless region; the problem is interpretive. I believe the main phase sequence is likely correct, but the abstract overstates the result in two ways: the 'only complex eigenspectra' statement is internally inconsistent, and the 'non-trivial semi-metallic' label is not backed by a valid invariant. Both issues are fixable within the scope of the manuscript. I do not have concerns about data fabrication or citation ethics."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a careful toy-model paper worth taking seriously, but it overreaches in the abstract and in one interpretive label.\n\nWhat is new: the four-site-per-cell SSH model with alternating Hermitian and non-Hermitian dimers is not in the cited literature. The authors derive closed-form critical values uc1, um, uc2, and those match the numerical band closings. The main sequence—non-trivial insulator → gapless → non-trivial insulator for θ=π/4, and trivial → gapless → non-trivial for θ=3π/4—is supported by the spectra and by the OBC edge-mode analysis. The IPR and disorder-robustness checks are a nice addition. This is reproducible, well-parameterized work.\n\nThe soft spots are real but mostly cosmetic. First, the abstract claims the system supports only complex eigenspectra for all u≠0, but the paper's own figures (e.g., Fig. 2(c)) show real-energy windows for intermediate u. That is a straightforward overclaim and should be fixed. Second, the 'non-trivial semi-metallic phase' label relies on the biorthogonal Zak phase from Eq. (16) evaluated across a region where bands touch at exceptional points; the eigenvector frame is not continuous, so the value is not a quantized invariant. The authors concede it 'fluctuates around 2π.' The insulator-gapless-insulator sequence does not depend on that label, but the 'unusual transition between non-trivial phases through non-trivial semi-metal' is exactly what that label is carrying. If the semi-metal cannot be assigned a well-defined non-trivial invariant, the transition becomes a gap-closing-reopening between two non-trivial insulators, which is less surprising. Third, 'weak non-Hermitian skin effect' is a stretch—no macroscopic mode localization is observed, only different IPR between left and right edge modes. That should be described as asymmetry, not NHSE. Fourth, the Fig. 8 caption says θ=π/4 where the text and content are θ=3π/4.\n\nNone of these are fatal. The central band structure and the phase sequence survive. This is the kind of paper a serious referee can fix with targeted revisions.\n\nWho it is for: people working on non-Hermitian topological models, especially SSH variants. It does not claim experimental realization, but it suggests photonic lattices. I would send it to peer review, with a request to rewrite the abstract, soften the semi-metal label, and reword the skin-effect claim.","headline":"A clean analytical band-structure study of a new non-Hermitian SSH variant; the core phase sequence is solid, but the abstract and the 'non-trivial semi-metal' label overreach.","tokens_in":15587,"tokens_out":2798,"would_cite":false,"duration_ms":24626,"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":"Strong loss-gain sends every SSH chain to the same topological phase.","keywords":["non-Hermitian SSH model","topological phase transition","exceptional points","Zak phase","semi-metallic phase","edge states","non-Hermitian skin effect","loss-gain strength"],"falsifier":"Compute a regularized topological invariant, such as a winding number along a path in $k$ that avoids the exceptional points, across the semi-metallic region: if the invariant does not remain pinned at the non-trivial value for all $u\\in(u_{c1}, u_{c2})$, the claim that the semi-metal is non-trivial—and with it the distinctive non-trivial-to-non-trivial story—fails, even though the insulator-gapless-insulator sequence would remain.","tokens_in":14588,"feed_emoji":"⚛️","tokens_out":6710,"duration_ms":48212,"temperature":0.7,"pith_summary":"This paper studies a non-Hermitian Su-Schrieffer-Heeger (SSH) chain in which alternating unit cells carry balanced loss and gain. It claims that strong non-Hermiticity $u$ drives every parameter regime into a non-trivial insulating phase: even if the chain is a trivial insulator in the Hermitian limit, raising the loss-gain strength forces a transition, through a gapless semi-metallic phase, to a non-trivial insulator. If the Hermitian limit is already non-trivial, the system leaves that phase, passes through a semi-metallic gapless region, and returns to a non-trivial insulator with different edge physics. The two non-trivial insulating phases are distinguished under open boundary conditions: one has two zero-energy edge modes, the other has one zero mode and one non-zero mode, and the model shows a weak non-Hermitian skin effect.","feed_headline":"Strong loss-gain sends every SSH chain to the same topological phase","feed_subtitle":"A gapless semi-metal mediates every route, even between two same-topology insulators.","key_machinery":"The central object is the four-band momentum-space Hamiltonian of Eq. (5), built from alternating Hermitian dimers ($c,d$) and non-Hermitian dimers ($a,b$) with onsite potentials $\\pm iu$. It obeys chiral symmetry $\\Gamma = \\tau_z \\otimes \\sigma_0$ and belongs to the BDI$^\\dagger$ symmetry class. The argument is carried by the trajectories of two exceptional points, EP1 and EP2, whose collision points define the critical loss-gain strengths $u_{c1}$, $u_m$, and $u_{c2}$; the region between $u_{c1}$ and $u_{c2}$ is gapless (semi-metallic) because of band merging at these exceptional points. Topological phases are classified with the Zak phase of Eq. (16) evaluated on biorthonormal left and right eigenstates.","core_discovery":"The central claim is that for loss-gain strength $u > u_{c2}$ the bulk always stabilizes in a non-trivial insulating phase, regardless of whether the Hermitian limit at $u=0$ is trivial or non-trivial. In the trivial case the transition runs trivial insulator $\\to$ gapless semi-metal $\\to$ non-trivial insulator; in the non-trivial case the transition runs non-trivial insulator $\\to$ (claimed non-trivial) semi-metal $\\to$ non-trivial insulator. Gap closing at exceptional points produces the intermediate semi-metallic region, and although gap closing usually signals a change in topology, here it mediates a return to the same topological phase. Under open boundaries the two non-trivial insulating phases differ: the low-$u$ phase hosts two zero-energy edge modes, the high-$u$ phase hosts a left zero mode and a right non-zero mode, with left modes more localized, indicating a weak non-Hermitian skin effect. The paper further reports that these edge states are robust to Gaussian disorder.","pith_inferences":["If the fluctuating Zak phase in the gapless region is replaced by an invariant that is well-defined despite the exceptional points, the 'non-trivial semi-metal' label may sharpen, but the robust part of the paper—strong loss-gain always ends in a non-trivial insulator—does not depend on that label.","The alternating Hermitian/non-Hermitian dimer pattern suggests a design principle: spatially periodic loss-gain can act as a topological 'reset' at strong coupling, which could be tested in other one-dimensional models such as the Kitaev chain or Aubry-André model.","The predicted asymmetry in inverse participation ratio between left and right edge modes is directly measurable in photonic waveguide arrays with engineered loss and gain; observing the asymmetry would verify the weak skin effect."],"forward_implications":["For $u > u_{c2}$, the bulk is always a non-trivial insulator for every dimerization angle $\\theta$, so a sufficiently strong loss-gain term can override the designed trivial or non-trivial character of the Hermitian chain.","The gapless semi-metallic phase can mediate a transition between two topologically equivalent insulating phases, not only between topologically distinct ones, which runs against the usual reading of gap closing as a topology-change indicator.","Under open boundary conditions, the low-$u$ and high-$u$ non-trivial insulators are not identical: they carry different zero/non-zero edge-mode structures, and the model exhibits a weak non-Hermitian skin effect with left modes more localized than right modes.","Bulk-boundary correspondence holds in the insulating regimes but is only weakly preserved near the boundaries of the gapless region, where non-zero discrete modes appear.","The edge states, including those in the gapless region, remain stable under Gaussian disorder."],"supporting_citations":[{"why":"Supplies the SSH model, the Zak phase definition, and the standard topological phase classification used as the Hermitian baseline.","marker":"[4]"},{"why":"Provides the non-Hermitian SSH framework and the Zak phase formula on biorthonormal eigenstates used to classify the phases.","marker":"[11]"},{"why":"Underpins the folded four-band spectrum, the PT-symmetric treatment, and the Zak phase analysis for the non-Hermitian SSH case.","marker":"[18]"},{"why":"Gives the symmetry classification that places the model in the BDI$^\\dagger$ class, justifying the expected phase-transition behavior.","marker":"[26]"},{"why":"Demonstrates that non-Hermiticity can induce a non-trivial insulating phase in a topologically trivial gapless system, the direct precedent for the paper's main result.","marker":"[31]"},{"why":"Provides the fully non-Hermitian SSH comparison showing no non-Hermitian skin effect, against which the paper's weak skin effect is contrasted.","marker":"[33]"},{"why":"States the expected role of semi-metallic phases in mediating topological switching, which the paper's non-trivial-to-non-trivial transition challenges.","marker":"[40]"}],"fun_headline_variants":["Loss-gain strength alone picks the final topological phase in SSH chains","SSH chains hit same topological phase no matter their start","Unusual route: non-trivial insulator to same non-trivial insulator","High loss-gain forces SSH chains into one topological phase","Semi-metal mediates a topological phase that never actually changes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that the gapless semi-metallic region is itself non-trivial rests on a Zak phase computed where bands touch at exceptional points, where a quantized invariant is not well-defined; the paper notes the value fluctuates around $2\\pi$.","fun_headline_variants_meta":{"raw":{"variants":["Loss-gain strength alone picks the final topological phase in SSH chains","SSH chains hit same topological phase no matter their start","Unusual route: non-trivial insulator to same non-trivial insulator","High loss-gain forces SSH chains into one topological phase","Semi-metal mediates a topological phase that never actually changes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000194,"raw_usage":{"total_tokens":1397,"prompt_tokens":1032,"completion_tokens":365,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":279}},"tokens_in":648,"tokens_out":365,"duration_ms":3903,"temperature":1.0,"reasoning_tokens":279,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:18:41.182694+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute a regularized topological invariant, such as a winding number along a path in $k$ that avoids the exceptional points, across the semi-metallic region: if the invariant does not remain pinned at the non-trivial value for all $u\\in(u_{c1}, u_{c2})$, the claim that the semi-metal is non-trivial—and with it the distinctive non-trivial-to-non-trivial story—fails, even though the insulator-gapless-insulator sequence would remain.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the SSH model, the Zak phase definition, and the standard topological phase classification used as the Hermitian baseline."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Underpins the folded four-band spectrum, the PT-symmetric treatment, and the Zak phase analysis for the non-Hermitian SSH case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the symmetry classification that places the model in the BDI$^\\dagger$ class, justifying the expected phase-transition behavior."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates that non-Hermiticity can induce a non-trivial insulating phase in a topologically trivial gapless system, the direct precedent for the paper's main result."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the fully non-Hermitian SSH comparison showing no non-Hermitian skin effect, against which the paper's weak skin effect is contrasted."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the expected role of semi-metallic phases in mediating topological switching, which the paper's non-trivial-to-non-trivial transition challenges."}],"review_version":1}