{"id":"1a62e27d-cc39-445f-a97a-e7551fecba3b","arxiv_id":"1908.08834","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Quadratic Lindbladians are captured by a non-Hermitian matrix Z in ten Bernard-LeClair symmetry classes, giving topologically protected zero-frequency edge modes with finite lifetimes.","lead":"This paper shows that a broad class of open fermionic systems, described by quadratic Lindblad master equations, falls into ten symmetry classes whose topology determines whether the system has stable, finite-lifetime edge modes. It gives a framework for predicting spectral features of dissipative topological systems, such as broadened zero-frequency peaks in Majorana wires.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exhaustiveness of the physical symmetry list (Eq. 6a-c) is asserted, not proven; the central ten-class claim depends on it.","rationale":"The paper's central claim is a new tenfold classification for quadratic Lindbladians, with the matrix Z = H + i Re[M] constrained to one of ten Bernard-LeClair classes. The proof structure is: (i) the Lindblad spectrum is determined by Z; (ii) Z must obey the spectral constraints Im λ ≤ 0 and {λ} = {-λ*}; (iii) these constraints leave only the three symmetry forms (6a-c); (iv) the real-line-gap classification of those ten classes matches the Altland-Zirnbauer table. Step (iii) is the least secure. The paper gives a per-symmetry eigenvalue argument for the standard BL relations, but does not systematically prove that no other relation can be compatible with complete positivity and the spectral constraints. Since step (iv) is only valid if the list of classes is exhaustive, this is a genuine proof gap, not a disagreement with consensus. The imported Ref. [14] table is standard and is not in itself a serious concern. The secondary issue in the Supplementary—the locality and symmetry preservation of the Cholesky-type deformation path—is real but does not affect the spectral classification itself. The numerical examples and the microscopic TRS derivation provide independent support for the plausibility of the ten classes, so the appropriate disposition remains conditional: the paper is likely correct, but the exhaustiveness step should be completed before the classification is fully established.","tokens_in":14561,"tokens_out":28691,"duration_ms":317896,"concrete_test":"Use the Bernard-LeClair classification of Ref. [19] to enumerate all relations of the forms U Z U† = ±Z, U Z^T U† = ±Z, U Z* U† = ±Z, U Z† U† = ±Z (and phase variants) for a matrix Z satisfying Z = -Z* and Im σ(Z) < 0. Compute the induced map on the eigenvalue multiset and retain only maps preserving the lower-half-plane condition and the {λ} = {-λ*} pairing. Verify that the surviving relations are exactly those implied by (6a-c); equivalently, check an ensemble of random 2x2 and 4x4 matrices Z = iR with R real and no real eigenvalues, confirming that no unitary satisfies any non-listed relation. Any counterexample would invalidate the exhaustiveness step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central classification statement is that, because the eigenvalues of Z obey Im λ ≤ 0 and {λ} = {-λ*}, the only admissible symmetry relations are (6a-c), so Z must belong to one of ten Bernard-LeClair classes. This is the load-bearing step that licenses importing the Ref. [14] real-line-gap table. The paper's argument for exhaustiveness is an eigenvalue-level dismissal: relations inducing λ → λ* or λ → -λ would create amplifying partners and are 'unphysical.' That rules out the familiar pseudo-Hermitian and transpose-PHS BL relations, but it is not a proof over the full BL classification of Ref. [19], nor does it treat mixed relations (e.g., with phase factors) or symmetries of the full Lindbladian (3) acting on the Y-block with no effect on Z. Without such a proof, the claim 'must belong to one of ten classes' is an assertion, and the topological classification inherits that gap. This concern is about completeness, not correctness of the examples; the Kitaev and SSH numerics are consistent with the stated classes.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies quadratic Lindbladians—Markovian open fermionic systems with quadratic Hamiltonians and linear dissipators—and reduces the Lindblad spectrum to the eigenvalues of a non-Hermitian single-particle matrix Z = H + i Re[M]. The central claim is that physical constraints on this spectrum, Im λ ≤ 0 and {λ} = {−λ*}, restrict Z to one of ten Bernard-LeClair symmetry classes generated by the three symmetries in Eq. (6). Under a real-line gap, these ten classes are asserted to yield a topological classification identical to the Altland-Zirnbauer tenfold way. The paper further argues that spectral topology is independent of steady-state topology by constructing a deformation to a system with the same spectrum but a trivial infinite-temperature steady state. The claims are illustrated with a dissipative Kitaev chain and an open SSH chain in the main text and Supplemental Material.","tokens_in":14757,"tokens_out":10279,"duration_ms":109234,"significance":"If the central claims hold, this is an important contribution: it extends the tenfold-way classification to a broad class of open fermionic systems, predicts robust zero-frequency edge excitations with finite lifetimes that should be visible as broadened spectroscopic peaks, and cleanly separates spectral from steady-state topology. The paper uses no fitted parameters, relies on the standard Prosen mapping, and checks its classification against the external Bernard-LeClair/Kawabata benchmark and against explicit numerical spectra. The examples are instructive and consistent with the proposed classes. The main weakness is that the exhaustiveness of the physical symmetry list in Eq. (6) is not proved, and the Supplemental Material's locality argument for the steady-state-independence deformation is asserted rather than rigorously justified.","major_comments":[{"comment":"The central classification claim—that Z must belong to one of the ten classes generated by Eqs. (6a–c)—is asserted rather than proved. The text rules out elementary relations λ→λ* and λ→−λ on the grounds that they create amplifying partners, but it does not systematically enumerate all Bernard-LeClair single-particle symmetry relations, including phase-factored and Hermitian-adjoint variants, nor does it address whether a symmetry of the full Lindbladian (3) that acts only on the Y block could add constraints or alter the class of Z. Because the real-line-gap table of Ref. [14] is imported precisely for the ten classes in Eq. (6), this exhaustiveness point is load-bearing for the claimed classification. I request either a proof that any symmetry of a physical quadratic Lindbladian compatible with Im λ ≤ 0 and {λ} = {−λ*} induces a relation of the form (6a–c), or a precise statement limiting the classification to the three symmetry forms considered.","section":"Non-Hermitian tenfold way, Eq. (6)"},{"comment":"The proof that spectral and steady-state properties are independent depends on the existence of a local, continuous family A(s) satisfying Eq. (S23), with the same locality properties as the original dissipators and with a continuous path of orthogonal rotations Q(s). The paper asserts that a standard Cholesky decomposition gives a solution A(s) with the same locality and that Q(s) can be chosen to make the deformation continuous; neither statement is proved. For banded matrices in one dimension the banded Cholesky factor is standard, but the higher-dimensional and positive-semidefinite cases, as well as the continuity of the chosen gauge, are not immediate. Since this deformation is the basis for the claimed decoupling of spectral and steady-state topology, the gap should be closed or the claim appropriately qualified.","section":"Supplemental Material, Eq. (S23)"}],"minor_comments":[{"comment":"There are small typos: 'femionic' in the abstract and 'irreversably' in the Supplemental Material.","section":"Abstract and Supplemental Material"},{"comment":"The superoperators c_j and c_j† act on the density matrix through formulas involving the fermion parity superoperator P_F; this object is referenced by a citation but not defined in the main text, which makes the presentation hard to follow for readers unfamiliar with the Prosen construction.","section":"Main text, Eq. (3)"},{"comment":"The statement that the symmetry-breaking dissipators (S4) force the SSH model into the trivial sector of the Z2 classification is not derived; a short argument showing that the relevant Z2 invariant evaluates to zero would make the example self-contained.","section":"Supplemental Material, dissipative SSH chain"},{"comment":"The distinction between spectral edge modes (excitations of the Lindbladian) and steady-state zero modes is stated explicitly only near Fig. 1; it would be helpful to restate this distinction when discussing the physical consequences of the edge modes in the introduction and outlook.","section":"Main text, Fig. 1 and outlook"}],"recommendation":"major_revision","confidential_remarks":"The exhaustiveness of Eq. (6) is the main obstacle to accepting the central claim as proven; the rest of the paper is clear and the examples are convincing. If the authors can provide a rigorous enumeration of admissible symmetries, or alternatively restrict the scope of the theorem accordingly, I would be inclined to support acceptance. The locality argument in the Supplemental Material also needs to be made rigorous."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this is a real contribution, worth your time. The paper gives a tenfold classification for quadratic Lindbladians—the open-system analogue of the Altland–Zirnbauer table—and shows that the full Lindblad spectrum, not just the steady state, can host topologically protected edge modes pinned to zero real frequency with finite lifetime. That is new and genuinely useful.\n\nThe paper does a lot right. The Prosen mapping is used cleanly: the Lindblad spectrum reduces to a non-Hermitian matrix Z = H + i Re[M], and the upper-triangular structure means Y drops out. The core symmetry argument is correct: the spectral constraints Im λ ≤ 0 and {λ} = {-λ*} forbid any pairing that sends a decaying mode to an amplifying one (λ → λ* or λ → -λ). That leaves exactly the three forms in Eq. (6): transpose TRS, conjugate PHS, and pseudo-anti-Hermiticity. I checked the logic; it holds up. The two examples are well chosen. The Kitaev chain with dissipators stays in class BDI and keeps Re λ_edge = 0, with a finite lifetime; the SSH case with symmetry-breaking dissipation drops to class D and the edge modes gap—nice demonstration of the sensitivity.\n\nThe soft spots are minor. The exhaustiveness of Eq. (6a-c) is asserted more than proven; a short lemma would settle it, and a referee should insist on that. The topological classification table is imported from Kawabata et al. [14] rather than rederived—standard practice, but it means the central result inherits that external classification. The supplement's proof that spectral and steady-state topology decouple is clever, but the locality argument for the deformation path is underdefended; it would be nice to see a rigorous statement. None of this shakes the main conclusion.\n\nCitation pattern is fine: self-citations are contextual, and the BL classification is cited properly. The numerics are minimal but sufficient as illustrations, not as fits.\n\nWho reads this? Anyone in non-Hermitian topology or open fermionic systems. It will likely become a standard reference. Send it to peer review; it deserves a serious referee, and I'd accept after minor revision—mainly the exhaustiveness lemma and a tightened locality claim.\n\nRegards.","headline":"A real and useful extension of the tenfold way to quadratic Lindbladians; the symmetry-list argument is terse but sound.","tokens_in":15257,"tokens_out":7877,"would_cite":true,"duration_ms":83151,"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":"Open fermionic systems with linear dissipation fall into the same ten symmetry classes as closed topological insulators, with edge modes pinned in frequency but finitely lived.","keywords":["quadratic Lindbladians","open quantum systems","non-Hermitian topological classification","Bernard-LeClair symmetry classes","Altland-Zirnbauer tenfold way","Majorana edge modes","dissipative Kitaev chain","steady-state topology"],"falsifier":"Construct a quadratic Lindbladian whose spectral matrix respects a symmetry outside the three listed forms, while the master equation remains completely positive and the real-line gap stays open; if its edge-mode count differs from the Altland-Zirnbauer tenfold table, the claimed exhaustion of symmetry types fails. A more direct experimental test: measure the zero-frequency spectroscopic peak of a dissipative Kitaev chain while increasing the strength of parity-breaking dissipators; if the peak centroid moves off zero without any symmetry change, the frequency-pinning claim fails.","tokens_in":14298,"feed_emoji":"⚛️","tokens_out":6240,"duration_ms":62649,"temperature":0.7,"pith_summary":"This paper establishes a topological classification for open fermionic systems whose density matrix evolves under a Lindblad master equation with a quadratic Hamiltonian and linear dissipators, called quadratic Lindbladians. It argues that the full excitation spectrum of such a system is controlled by a single non-Hermitian matrix $Z = H + i\\mathrm{Re}[M]$, and that the physical constraints of decay and Hermiticity force $Z$ to belong to exactly one of ten non-Hermitian Bernard-LeClair symmetry classes. Under a real-frequency gap, those ten classes carry the same topological classification as the Altland-Zirnbauer tenfold way for closed systems. As a result, a topologically nontrivial Lindbladian has gapless edge excitations whose phase-oscillation frequencies are pinned inside the gap but which generically decay with finite lifetimes, and these spectral features are independent of the steady state. This matters because it tells which dissipative perturbations preserve the edge modes of topological materials and how those modes appear in spectroscopy.","feed_headline":"Open quantum systems get a tenfold way too","feed_subtitle":"Linear dissipation still leaves topological edge modes protected at zero frequency, only now they decay.","key_machinery":"The load-bearing object is the non-Hermitian single-particle spectral matrix $Z = H + i\\mathrm{Re}[M]$, whose $2N$ eigenvalues give the complex Lindblad spectrum through the quasiparticle decomposition $\\mathcal L = 4\\sum_j \\lambda_j \\bar\\beta_j^\\dagger \\beta_j$. Prosen's triangularization puts the Lindbladian in block upper-triangular form with $Z$ as the diagonal block, so the imaginary part of $M$ drops out of the spectrum. The symmetry argument rests on the restricted list of allowed symmetry actions: transposition-based time-reversal $Z = U_T Z^T U_T^\\dagger$, conjugation-based particle-hole $Z = -U_C Z^* U_C^\\dagger$, and pseudo-anti-Hermiticity $Z = -U_S Z^\\dagger U_S^\\dagger$. These are the only actions compatible with the constraints $\\mathrm{Im}[\\lambda_i]\\le 0$ and spectral pairing $\\{\\lambda\\} = \\{-\\lambda^*\\}$; they generate ten Bernard-LeClair classes, and the real-line-gap classification table for those classes is the same as the Altland-Zirnbauer tenfold way.","core_discovery":"The central discovery is that the Lindblad spectrum of any quadratic Lindbladian is fully determined by the eigenvalues of a non-Hermitian single-particle matrix $Z = H + i\\mathrm{Re}[M]$, where $H$ is the first-quantized Hamiltonian and $M$ is built from the dissipators. Because eigenvalues must lie in the lower half-plane and come in anti-complex-conjugate pairs, the only symmetry operations $Z$ can respect are transposition-based time-reversal, conjugation-based particle-hole, and pseudo-anti-Hermiticity; combinations of these generate exactly ten Bernard-LeClair symmetry classes, which reduce to the ten Altland-Zirnbauer classes when dissipation vanishes. Under a real-line gap condition, the topological classification of these ten classes coincides with the conventional tenfold way. Consequently a topologically nontrivial Lindbladian supports robust edge modes pinned to zero frequency, with generically finite decay rates, and the existence of such spectral edge modes carries no implication for the topological character of the steady state; the paper proves this independence by constructing a continuous deformation to an auxiliary Lindbladian with identical spectrum and symmetries but a trivial infinite-temperature steady state.","pith_inferences":["The restriction of admissible symmetries to the three forms in Eq. (6a-c) comes from spectral decay and Hermiticity constraints, so the same ten-class structure may survive in interacting or non-quadratic open fermion systems that respect those constraints.","The classification suggests that random dissipative systems should display ten universality classes in their complex spectral statistics, mirroring the Altland-Zirnbauer random-matrix classes; this is testable in open quantum dots and disordered wires.","Because only the real frequency of an edge mode is protected while its decay rate is not, dissipative engineering could tune edge-mode lifetimes without closing the spectral gap, which may be useful for controlling quasiparticle poisoning.","The proved independence of spectral and steady-state topology implies that spectroscopic probes alone cannot certify steady-state topology in open systems; correlation-function or steady-state-response measurements are needed."],"forward_implications":["A symmetry-preserving dissipative Kitaev chain remains in class BDI, so its Majorana edge modes stay pinned at zero frequency but acquire a finite lifetime through quasiparticle poisoning.","Spectroscopic signatures of topological edge modes in open systems are broadened peaks centered inside the gap; the peak center is protected, while the width is set by the dissipative coupling.","Spectral topology and steady-state topology are independent: robust in-gap edge modes imply nothing about the steady-state density matrix, and a trivial steady state can coexist with a nontrivial spectrum.","Adding Hamiltonian dynamics to a purely dissipative topological system generically gives its desired in-gap edge modes a finite lifetime, making purely dissipative steady-state edge modes fragile.","Protected imaginary-gap edge modes cannot exist for quadratic Lindbladians because the imaginary part of the Lindblad spectrum is constrained to be non-positive."],"supporting_citations":[{"why":"Supplies the quadratic-Lindbladian diagonalization: the Lindblad spectrum is obtained by diagonalizing a non-Hermitian fermionic superconductor in Bogoliubov-de Gennes form.","marker":"[17, 18]"},{"why":"Provides the Bernard-LeClair symmetry classification of non-Hermitian matrices under a real-line gap, which the paper maps onto its ten Lindbladian classes.","marker":"[14]"},{"why":"Defines the Bernard-LeClair non-Hermitian symmetry classes from which the ten Lindbladian classes are drawn.","marker":"[19]"},{"why":"Defines the Altland-Zirnbauer tenfold classification of closed Hermitian fermionic systems that the Lindblad classification reduces to in the closed limit.","marker":"[29]"},{"why":"Establishes the conventional tenfold way for Hermitian free-fermion models that the paper extends to quadratic Lindbladians.","marker":"[30, 31]"},{"why":"Provides the prior tenfold-way classifications for purely dissipative systems based on steady-state properties, the contrast point for the paper's spectral and steady-state independence claim.","marker":"[20, 21]"},{"why":"Previous study of a dissipative Kitaev chain whose parity-breaking dissipators give the finite-lifetime Majorana modes discussed in the paper's example.","marker":"[28]"},{"why":"Provides the Kitaev chain model in class BDI whose Majorana edge modes are tested under linear dissipators.","marker":"[34]"},{"why":"Contains the proof of spectral and steady-state independence via a deformation to a trivial steady state, and the microscopic derivation of the non-Hermitian time-reversal symmetry.","marker":"[32]"}],"fun_headline_variants":["Tenfold way extends to open fermionic systems","Quadratic Lindbladians reveal a tenfold symmetry class","Topological edge modes survive with finite lifetimes","Tenfold way for open systems: dissipation doesn't kill topology","Bernard-LeClair classes give open systems a tenfold way"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole ten-class scheme depends on the assertion that no symmetry outside the three listed forms can act on the spectral matrix while keeping the density-matrix evolution physical; that exhaustiveness is stated rather than proved.","fun_headline_variants_meta":{"raw":{"variants":["Tenfold way extends to open fermionic systems","Quadratic Lindbladians reveal a tenfold symmetry class","Topological edge modes survive with finite lifetimes","Tenfold way for open systems: dissipation doesn't kill topology","Bernard-LeClair classes give open systems a tenfold way"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000423,"raw_usage":{"total_tokens":2159,"prompt_tokens":917,"completion_tokens":1242,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":533,"completion_tokens_details":{"reasoning_tokens":1163}},"tokens_in":533,"tokens_out":1242,"duration_ms":9661,"temperature":1.0,"reasoning_tokens":1163,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:28:07.605687+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a quadratic Lindbladian whose spectral matrix respects a symmetry outside the three listed forms, while the master equation remains completely positive and the real-line gap stays open; if its edge-mode count differs from the Altland-Zirnbauer tenfold table, the claimed exhaustion of symmetry types fails. A more direct experimental test: measure the zero-frequency spectroscopic peak of a dissipative Kitaev chain while increasing the strength of parity-breaking dissipators; if the peak centroid moves off zero without any symmetry change, the frequency-pinning claim fails.","supporting_citations":[{"cited_title":"A classification of non- hermitian random matrices,","cited_arxiv_id":null,"evidence_quote":"Defines the Bernard-LeClair non-Hermitian symmetry classes from which the ten Lindbladian classes are drawn."},{"cited_title":"van Caspel, S","cited_arxiv_id":null,"evidence_quote":"Previous study of a dissipative Kitaev chain whose parity-breaking dissipators give the finite-lifetime Majorana modes discussed in the paper's example."},{"cited_title":"Contains Refs","cited_arxiv_id":null,"evidence_quote":"Contains the proof of spectral and steady-state independence via a deformation to a trivial steady state, and the microscopic derivation of the non-Hermitian time-reversal symmetry."}],"review_version":1}