{"id":"b296f199-b68d-45ec-bc75-a9163b9cd80b","arxiv_id":"2607.22976","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The non-Hermitian skin effect at domain walls is caused by relative spectral winding between adjacent domains, and the ring spectrum splits into standing-wave and flux-sensitive traveling-wave sectors with distinct generalized Brillouin zone conditions.","lead":"This paper explains why waves bunch up at the seams when a ring is made of several different lattice segments: it happens only when the two neighboring segments twist around the complex energy in different ways. It also predicts a new family of traveling waves, unique to such multi-segment rings, that respond to a loop of magnetic flux.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Dominant-term reduction in SIII assumes leading coefficients A_{w;I}(E) never vanish along predicted spectral arcs; if they cancel, Case I/II GBZ conditions miss states — reader's flagged caveat, worth one explicit check.","rationale":"The central claim has two parts: (i) the relative-winding criterion, which is supported by the Toeplitz-index proof and is robust; and (ii) the GBZ classification via determinant expansion, which contains the soft spot. The determinant expansion itself is exact, but the thermodynamic reduction to Case I/Case II depends on the dominance of the largest exponential orders and on non-cancellation of their coefficients. The authors explicitly limit the theory by excluding discrete boundary states, and they do not prove that the coefficients A_{w;I}(E) remain nonzero on the spectral arcs. This is a genuine gap in the general proof. It does not invalidate the paper because the authors flag the limitation, the one detailed example matches direct diagonalization, and generic algebraic functions vanish only at isolated points unless structural zeros are present. Still, because the Case II traveling-wave sector is the headline novelty and is obtained from this reduction, the nonvanishing condition should be checked rather than assumed. No other concern appears more load-bearing: the Toeplitz-index argument is standard, the combinatorial determinant count is internally consistent, and the flux-winding argument follows once the GBZ classification is accepted.","tokens_in":28883,"tokens_out":14380,"duration_ms":164717,"concrete_test":"For the 3-DW model of Figs. 2–4, construct M(E) from Eq. (S57) and numerically evaluate the coefficients A_{w;I_1,I_2,I_3}(E) on a fine grid along the Ronkin-predicted Case I and Case II curves. Verify that for each maximal sector w, the sum of coefficients of the terms attaining the maximal exponential order is nonzero on every curve segment. A stronger direct check: for N=(58,41,87) and larger scaled sizes, compute |det M(E_N)| for E_N on a predicted spectral curve; if |det M|/exp(N Λ_*) tends to zero systematically as N grows, the dominance reduction is invalid; if it remains O(1), the concern is cleared.","verdict_should_be":"UNCHANGED","load_bearing_attack":"SM SIII expands det M(E) = Σ_{w,I} A_{w;I}(E) ∏_α (∏_{m∈I_α} β_{α,m})^{N_α} (Eq. S61) and then reduces to the two GBZ conditions (Eqs. S74–S75) by keeping only the largest modulus in the exponential-order set ζ. This reduction is valid only if, on every arc of the predicted continuum spectrum, the coefficients of the maximal-order terms are nonvanishing and do not cancel identically. The paper explicitly excludes the case of a single vanishing dominant coefficient — 'discrete boundary states' (SM SIII, SI) — but it does not exclude a more dangerous scenario: an identically zero leading coefficient, or an identically zero sum of all maximal-order coefficients, over an open energy arc. Since A_{w;I}(E) are determinants of submatrices of the interface matrices L_α, G_α, they are model-dependent algebraic functions; no general argument or model-specific check is supplied. If such cancellation occurred, the true dominant sector would shift and the predicted Case I/Case II spectrum would miss a continuum of states. The numerical agreement for the single 3-DW example (Figs. 3, S1, S2) is reassuring but does not prove the general statement. This is essentially the same load-bearing assumption the reader identified, made more precise.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies one-dimensional non-Hermitian systems arranged as a ring of n translationally invariant domains. Its central claims are: (i) an eigenstate localized at the interface α|α+1 exists only if the relative point-gap winding Δα(E)=w_{α+1}(E)-w_α(E) is positive (right eigenstates) or negative (left eigenstates), making the NHSE at domain walls topologically protected; (ii) in the thermodynamic limit the continuum eigenspectrum of the n-DW ring is characterized by two GBZ conditions—Case I (standing-wave-like: an equal-modulus pair inside a maximal common-winding sector) and Case II (traveling-wave-like: Σ r_α μ_{α,m_α(w)}=0); (iii) the Case II sector is unique to closed domain-wall rings and carries a nonzero flux spectral winding. The authors support (i) by eigenmode counting and a Toeplitz-index proof, and (ii) by a boundary-condition determinant expansion and, independently, by a constrained Ronkin-function calculation; numerical diagonalization of a 3-domain model is used to check the results.","tokens_in":29102,"tokens_out":14417,"duration_ms":140655,"significance":"If the claims hold, this is an important extension of non-Bloch band theory beyond translationally invariant single-bulk systems. The relative-winding criterion is a clean, parameter-free topological statement with a rigorous-looking proof, and the predicted traveling-wave sector—distinguished by a global round-trip condition and flux spectral winding—is a concrete, falsifiable new phenomenon. The agreement between two independent derivations (determinant expansion and Ronkin function) and the direct numerical comparison are strengths. However, because the dominant-term reduction in the determinant expansion rests on an unproven nonvanishing assumption for the coefficients A_{w;I}(E), the claimed completeness of the Case I/II characterization is not yet fully established. The paper has no fitted parameters and the numerical work is reproducible in principle from the stated model.","major_comments":[{"comment":"SM SIII, Eqs. (S61)-(S75): the reduction to Case I/II assumes the maximal-order terms in det M(E) are present with nonzero coefficients. The text excludes a single vanishing dominant coefficient as a discrete boundary state (SM SIII after Eq. (S62); SM SI), but never excludes an identically zero leading coefficient, or an identically zero sum of all maximal-order coefficients, along an open energy arc. Since A_{w;I}(E) are determinants of submatrices of L_α and G_α, no general argument prevents such a cancellation; if it occurred, the true dominant sector would shift and the predicted Case I/II spectrum would miss a continuum. The G_1=0 limit (SM SV) shows coefficient structure can change the spectral condition. The single 3-DW numerical example is not a proof. Please prove nonvanishing on the predicted arcs or state the result under a generic-coupling assumption and characterize excepti","section":"SM SIII, Eqs. (S61)-(S75)"},{"comment":"SM SIII, 'Derivation of boundary equations': the bulk solution (S41) assumes nondegenerate roots of f_α(β;E)=0 and uses the eigenvector basis u_{α,m}. The GBZ condition itself, however, is an equal-modulus condition μ_{α,m}=μ_{α,m+1}; at a genuine root degeneracy β_{α,m}=β_{α,m+1} (an exceptional point), that basis collapses and Eqs. (S41)-(S61) require a Jordan-chain or limiting argument. This is not supplied. If the theory is meant to cover only nondegenerate characteristic roots, that restriction should be stated where Case I/II are introduced; otherwise the determinant expansion is incomplete at the very points it is meant to characterize.","section":"SM SIII, 'Derivation of boundary equations'"}],"minor_comments":[{"comment":"'Complete characterization' and 'fully resolves' overstate the scope given the acknowledged exclusion of discrete boundary states and the generic-coefficient caveat in SM SIII. Please qualify.","section":"Abstract/Discussion"},{"comment":"The unshaded regions are not explained; state whether Δ1(E) is nonpositive there or simply not shown.","section":"Fig. 2"},{"comment":"Specify how W_DW=1 is evaluated (for example, numerical quadrature of Eq. (12) over the flux with a stated grid). As written, the claimed value is not independently checkable.","section":"Fig. 4 / Eq. (12)"},{"comment":"The heuristic counting sentence equating s_α and d_α−s_α with hopping ranges for multiband models is informal; since the Toeplitz proof is rigorous, label the counting argument as a single-band illustration.","section":"SM SII"},{"comment":"The symbols w_- and w_+ are used before their definitions; define them at first use.","section":"SM SIII, Eq. (S62)"}],"recommendation":"major_revision","confidential_remarks":"The dominant-term concern is the main obstacle; the paper should not be accepted until it is resolved with a genericity statement or proof. The traveling-wave prediction and the relative-winding criterion are otherwise compelling and fit the journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know: this is a genuinely new theoretical result, not a repackaging. It generalizes the two-domain phenomenological DW analyses to n-domain rings, proves that skin localization is controlled by relative spectral winding, and constructs a full non-Bloch band theory through a constrained Ronkin function. The traveling-wave-like skin-mode sector is a credible new class, with its own GBZ condition and finite flux spectral winding. That is the core intellectual contribution.\n\nWhat it does well: the two independent derivations agree. The relative-winding criterion is proved via Toeplitz index theory and via eigenmode counting; the GBZ reduction comes from a detailed determinant expansion of the boundary-condition matrix. Both routes lead to the same Case I / Case II conditions. Numerical checks are consistent: spectrum, DOS, and domain-resolved GBZ match direct diagonalization for the one 3-DW example shown. The paper is honest about what it excludes: discrete boundary states are explicitly carved out. No parameter fitting anywhere, so the circularity burden is genuinely low.\n\nNow the soft spots. The stress-test note points to a real gap: the thermodynamic reduction in SM SIII keeps only the largest-modulus exponential terms and assumes the corresponding coefficients A_{w;I}(E) are nonvanishing on every predicted spectral arc. The authors exclude a single vanishing dominant coefficient (discrete boundary states), but they do not address an identically zero leading coefficient — or an identically zero sum of maximal-order coefficients — over an open energy arc. If such a cancellation happened, the Case I/II spectrum would miss a continuum of states. That is a precise, checkable assumption, and the paper does not supply a general argument or a second model where it is verified. I do not think it sinks the central claim — the numerical agreement for the single example is reassuring, and the structure of the determinant expansion makes exact cancellations look exceptional — but it should be named in any referee report. The other caveats are minor: only one three-domain example, no code/data artifacts, and the flux-winding argument in SVI is leading-order heuristic. The novelty claim about prior DW work being limited to two domains is plausible based on the cited literature, though I would not swear to the absence of an obscure paper somewhere.\n\nWho is this for: anyone working on non-Hermitian skin effects, non-Bloch band theory, or domain-wall systems. It deserves a serious referee, not a desk reject. The referee should push specifically on the coefficient-nonvanishing condition and ask for either a proof in a natural class of models or a second nontrivial example. My own verdict would be positive with that caveat flagged.","headline":"Solid theory paper that upgrades domain-wall NHSE to n-domain rings and finds a new traveling-wave skin sector; the main caveat is a coefficient-nonvanishing assumption in the thermodynamic reduction.","tokens_in":29710,"tokens_out":1268,"would_cite":true,"duration_ms":15814,"reading_group":"yes","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 domain-wall ring of non-Hermitian lattices, every skin mode is tied to a winding mismatch between adjacent domains, and the full spectrum is fixed by two generalized Brillouin zone conditions.","keywords":["non-Hermitian skin effect","domain-wall ring","spectral winding","generalized Brillouin zone","Ronkin function","traveling-wave skin modes","non-Bloch band theory","flux spectral winding"],"falsifier":"Take a three-domain ring and tune the parameters so that, at some energy E0, the relative winding Δ1(E0)>0 while the weighted sum Σ_α r_α μ_{α,m_α(w)} = 0 has no solution in the relevant sector; if a mode localizing at interface 1|2 nevertheless appears at E0, the necessity of the Case I/II conditions would be falsified. Alternatively, find a domain-wall ring whose diagonalized spectrum contains an eigenenergy not obeying either Case I or Case II.","tokens_in":28677,"feed_emoji":"","tokens_out":5022,"duration_ms":48041,"temperature":0.7,"pith_summary":"The paper establishes that the non-Hermitian skin effect in a one-dimensional domain-wall ring—several different non-reciprocal lattices joined in a closed loop—is governed by the difference in spectral winding between adjacent domains, not by any single domain's winding alone. It then develops a non-Bloch band theory, extending the Ronkin-function formalism to multi-domain rings, and derives the generalized Brillouin zone (GBZ) conditions that determine the thermodynamic eigenspectrum. The spectrum splits into two sectors: standing-wave-like skin modes, which are inherited from the conventional open-boundary GBZ of individual domains, and traveling-wave-like skin modes, which exist only because of the global round-trip condition of the ring. The traveling-wave sector carries a finite flux spectral winding, so it is boundary-sensitive and disappears when the ring is cut into an open chain. If correct, the theory fully predicts where skin modes localize and which energies appear in a domain-wall ring.","feed_headline":"Winding mismatch between adjacent domains sets where skin modes live","feed_subtitle":"A new theory shows the ring spectrum splits into standing-wave and traveling-wave sectors; the traveling sector vanishes if the ring is cut.","key_machinery":"The relative spectral winding Δα(E)=w_{α+1}(E)−w_α(E) between adjacent domains is the topological order parameter; its sign predicts the localization side of skin modes. The GBZ conditions are obtained from the constrained Ronkin function R(μ;E)=Σ_{α=1}^{n−1} r_α R_α(μ_α;E)+r_n R_n(−Σ r_α/r_n μ_α;E), whose flat regions correspond to imaginary-gauge transformations that equalize all domain windings; the collapse of these flat regions yields the two spectral cases. Case I generalizes the single-domain GBZ/aGBZ condition; Case II is a new round-trip resonance condition, λ_w(E)=Σ_α r_α μ_{α,s_α+w}(E)=0.","core_discovery":"The central claim is that an eigenstate of an n-domain-wall ring that is exponentially localized near the interface between domains α and α+1 exists only when the relative spectral winding Δα(E)=w_{α+1}(E)−w_α(E) is positive (for right eigenstates); a vanishing mismatch across every interface rules out point-gap-protected skin effects. The full thermodynamic spectrum is then fixed by two GBZ conditions derived from the collapse of flat regions of a constrained Ronkin function: Case I (standing-wave-like) requires an equal-modulus pair μ_{α,m(w)} = μ_{α,m(w)+1} within some domain, consistent with the weighted global constraint Σ_α r_α μ_{α,m} ≤ 0 ≤ Σ_α r_α μ_{α,m+1}; Case II (traveling-wave-l","pith_inferences":["The Case II traveling-wave condition Σ_α r_α μ_{α,m}=0 is essentially a non-Hermitian analogue of a quantized Aharonov-Bohm phase around the ring; one might test the theory by measuring the flux-induced spectral flow in an electric-circuit or photonic quantum-walk implementation.","The constrained-Ronkin collapse method may generalize to higher-dimensional domain-wall configurations or to disordered/quasiperiodic domain walls, where the same relative-winding logic could hold.","The explicit exclusion of discrete boundary states (where a single dominant coefficient vanishes) suggests a complementary theory is needed for those states; the paper's continuum spectrum may fail to capture edge-localized modes in finite rings.","The claim that standing-wave modes carry no flux spectral winding could be tested directly in a finite ring: the flux winding should jump exactly when a traveling-wave branch crosses the reference energy."],"forward_implications":["Skin-mode localization in a domain-wall ring is determined by the sign of the relative winding across each interface; a positive mismatch is necessary for an eigenstate to accumulate there.","The thermodynamic ring spectrum is exactly the union of the spectra from the two GBZ conditions, so it can be computed from the characteristic equations of the constituent domains without diagonalizing the full ring.","The traveling-wave-like sector is a distinct class of skin mode with no single-domain open-boundary analogue; it disappears under open boundary conditions, leaving only standing-wave-like modes.","The flux spectral winding is carried entirely by the traveling-wave sector, so a 2π flux insertion shifts these states' energies but not the standing-wave sector's.","Open domain-wall chains have no traveling-wave sector; their thermodynamic spectrum equals the union of the open-boundary spectra of the constituent domains."],"fun_headline_variants":["Skin modes live where spectral winding jumps between domains","Traveling skin modes vanish when the ring is cut","Two skin sectors emerge from winding mismatch in domain-wall rings","Winding mismatch dictates interface localization of eigenstates","New skin modes travel across domains, vanish on open boundaries"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The argument that the thermodynamic spectrum is fully captured by the two GBZ conditions assumes that, for every energy on the continuum spectrum, the coefficients of the exponentially dominant terms in the boundary-determinant expansion do not vanish, so the largest-modulus terms are never cancelled by accidental degeneracies or coefficient zeros.","fun_headline_variants_meta":{"raw":{"variants":["Skin modes live where spectral winding jumps between domains","Traveling skin modes vanish when the ring is cut","Two skin sectors emerge from winding mismatch in domain-wall rings","Winding mismatch dictates interface localization of eigenstates","New skin modes travel across domains, vanish on open boundaries"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000211,"raw_usage":{"total_tokens":1248,"prompt_tokens":739,"completion_tokens":509,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":483,"completion_tokens_details":{"reasoning_tokens":433}},"tokens_in":483,"tokens_out":509,"duration_ms":5698,"temperature":1.0,"reasoning_tokens":433,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T04:00:09.424159+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a three-domain ring and tune the parameters so that, at some energy E0, the relative winding Δ1(E0)>0 while the weighted sum Σ_α r_α μ_{α,m_α(w)} = 0 has no solution in the relevant sector; if a mode localizing at interface 1|2 nevertheless appears at E0, the necessity of the Case I/II conditions would be falsified. Alternatively, find a domain-wall ring whose diagonalized spectrum contains an eigenenergy not obeying either Case I or Case II.","supporting_citations":[],"review_version":1}