{"id":"43e5744c-a042-42d8-9780-4b1d30342fd2","arxiv_id":"2508.07448","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For Lindblad quantum spin systems on infinite lattices, the paper proves that thermodynamic and long-time limits defining NESS and TANESS commute under spectral-gap plus condition-number assumptions, and gives a model where spectral gaps alone fail.","lead":"The paper defines nonequilibrium steady states and their time averages for open quantum spin systems on infinite lattices, then proves when the thermodynamic limit and the infinite-time limit can be swapped. A worked model shows the two limits can disagree even when every finite subsystem has a uniform spectral gap, so the new condition number criterion is necessary.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uniform condition-number bound is load-bearing, but §5.4's check that it fails in the example is incomplete because V is non-unique; the conclusion is recoverable from noncommutation plus Theorem 8.","rationale":"The reader's weakest-assumption analysis points to the uniform condition-number bound, and that is indeed the assumption most directly responsible for the Λ-independent exponential decay in Theorem 8. The proof of Theorems 8 and 10 is structurally sound: it uses (A1)/(B1) for semisimple eigenvalues and (A2)/(B2) for the Jordan blocks, and the bound κ e^{-min{Δ,ϵ}t} is what lets the long-time limit pass through the thermodynamic limit. The central theorem is not threatened by the non-uniqueness of V, provided (A2) is read as an existential condition. The actual soft spot is the example: Section 5.4 lower-bounds the condition number of one particular Jordanizing map. Because V^ϵΛn is non-unique, this does not by itself prove that no uniformly bounded choice exists, and the paper does not explicitly invoke the noncommutation calculation plus Theorem 8 to close the gap. That is a presentation/verification gap in an auxiliary claim, not a fatal flaw in the main result. The reader's CONDITIONAL verdict is appropriate; my read does not move it. The concrete test of computing the minimal condition number over all Jordanizing maps for the §5 model would settle whether the example genuinely violates (A2) and would also confirm whether the exponential lower bound in §5.4 is an artifact of the chosen V or reflects an unavoidable divergence. I therefore agree with the reader's identification and recommend keeping the verdict unchanged pending that clarification.","tokens_in":30235,"tokens_out":21366,"duration_ms":224134,"concrete_test":"For the restriction Mn of LΛn to one symmetry sector (a matrix of size 2n), compute the minimal condition number κ_n^* = inf { κ(V) : V^{-1} Mn V is in Jordan form }, for n = 1,...,20, by optimizing over the similarity orbit of the Jordan form (or analytically using the family of bases {e1, e1−e2, ..., e1−e_{2n}} with column scaling). If κ_n^* diverges, the §5.4 conclusion survives despite the non-optimal V; if κ_n^* stays bounded, then the model would satisfy (A2), and Theorem 8 would predict commutation, contradicting the exact 0 vs 1/2 result in §5.2, requiring a re-examination of either the section or the theorem.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assumption is the uniform condition-number bound κΛ(V^ϵΛ) ≤ κ in (A2)/(B2). The paper itself notes (paragraph after Theorem 8) that the Jordanizing map V^ϵΛ is non-unique: V ↦ A∘V with A commuting with the Jordan form changes κΛ. Theorem 8 is valid if one reads (A2) existentially. But Section 5.4 computes a lower bound for one explicitly constructed V^ϵΛn; because a different V can have a much smaller condition number (e.g., by rescaling generalized eigenvectors), this lower bound does not by itself prove that no bounded choice exists, so it does not directly establish failure of (A2)/(B2). The advertised conclusion can be recovered from the exact noncommutation in §5.2 (1/2 vs 0) together with Theorem 8, but the paper does not state that argument in §5.4. Thus the gap is in the direct verification of the example, not in the proof of the sufficient condition.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Lindblad dynamics for quantum spin systems on infinite lattices. It defines NESS and TANESS as weak-* cluster points of long-time limits and long-time averages of an initial state evolved under the infinite-system dynamics, and observes that the thermodynamic limit and the long-time limit do not generally commute. The main results, Theorems 8 and 10, give sufficient conditions for commutativity: uniform finite-volume line/point spectral gaps, a uniform exceptional gap, and a uniform upper bound on the condition number of a Jordanizing map for the normalized Liouvillian. Under these assumptions the paper proves that the thermodynamic limit of finite-system NESS/TANESS exists, is independent of subnet choices, and equals the unique infinite-system NESS/TANESS for any initial state. The proof uses a finite-volume spectral projection and uniform exponential decay controlled by the spectral gaps and the condition number. The paper then constructs an explicit finite-range model on Γ=N for which the spectral gaps are uniformly nonvanishing (Δ=Δ_p=1/2, Δ_ex=1) but the two limits do not commute: for an explicitly chosen initial state, the infinite-system expectation of a local observable converges to 1/2 in the long-time limit, whereas the thermodynamic limit of finite-system long-time limits is 0. The noncommutation is established by exact computation, and the paper attributes the failure to the unboundedness of the condition number.","tokens_in":30476,"tokens_out":5192,"duration_ms":54924,"significance":"If correct, the paper provides a useful operator-algebraic framework for NESSs of infinite Lindblad systems and a concrete, checkable sufficient condition for when finite-volume steady states converge to infinite-volume ones. The result is nonempty: it covers systems with finite-volume Liouvillians that are sufficiently close to normal, uniformly in the volume, and it sharpens the known fact that spectral gaps alone are insufficient. The example is valuable: it is exactly solvable, with explicit spectra, generalized eigenspaces, and a transparent noncommutation mechanism. A notable strength is that the main theorems are proved in detail and the crucial finite-volume estimates are written out; the example's expectation values are computed explicitly rather than argued heuristically. The paper also honestly acknowledges the non-uniqueness of the Jordanizing map after Theorem 8, although, as discussed below, Section 5.4 does not fully follow through on that acknowledgment.","major_comments":[{"comment":"The direct verification that assumptions (A2)/(B2) fail is incomplete. The paper itself notes after Theorem 8 that V^ϵ_Λ is non-unique: replacing V^ϵ_Λ by A∘V^ϵ_Λ, where A commutes with the Jordan form, changes the condition number. Therefore the lower bound κ_{Λ_n}(V^ϵ_{Λ_n}) ≥ (1/2)(1−ϵ)^{2n−2} for the particular V^ϵ_{Λ_n} constructed in Eq. (38) does not prove that no Jordanizing map with uniformly bounded condition number exists. Since (A2)/(B2) are existential conditions, this is a genuine gap. The advertised conclusion is recoverable: the exact noncommutation in §5.2 (1/2 versus 0) together with Theorem 8 and Corollary 17 (uniform gaps) implies by contradiction that no uniformly bounded V^ϵ_Λ can exist. This argument should be stated explicitly in §5.4, and the text should distinguish the lower bound for the constructed V from the nonexistence of any bounded choice.","section":"§5.4 (concluding inequality before the final paragraph)"},{"comment":"In the non-semisimple case the displayed equation writes \"e^{t Re λ_h} e^{(∆−ϵ)t∥N_h∥}\" but the preceding line uses Re λ_h ≤ −∆ex_Λ ≤ −∆ex. The exponent should be ∆ex, not ∆; otherwise the inequality e^{t Re λ_h} e^{(∆−ϵ)t} ≤ e^{−ϵt} is not justified. The final e^{−ϵt} decay is correct once ∆ex is used, so this is a local typo, but it should be fixed.","section":"§4.3, proof of Theorem 8, estimate after Eq. (16)"}],"minor_comments":[{"comment":"The text says \"a subnet of (t)_{t>0}\" for the TANESS; it should be \"a subnet of (T)_{T>0}\" or, equivalently, of the net indexed by T.","section":"Definition 6"},{"comment":"Typo: \"witten\" should be \"written\".","section":"Appendix A, Definition 19"},{"comment":"The notation for the initial state ω0 specifies expectations only on the operators D_{2j−1}, D_{2j}; it would help the reader to note that this extends to a genuine state (e.g., a product state with odd sites in |↑⟩ and even sites maximally mixed), since the subsequent computations use evaluation on combinations of the D_m.","section":"Section 5.2, discussion after Eq. (24)"},{"comment":"Several minor typos: \"complex plain\" (p. 10), \"parpicular\" (p. 16), \"fields\" formatting issues in equations. These do not affect the mathematics but should be cleaned up.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's central sufficient-condition theorem appears sound and well proved. The main weakness is Section 5.4's incomplete proof of unboundedness of the condition number; however, the conclusion can be recovered from the already-established noncommutation and the main theorem. I therefore view this as a fixable load-bearing gap rather than a fatal flaw. The paper would be strengthened by making the existential reading of (A2)/(B2) explicit and by adding the short contradiction argument in Section 5.4. Fit with math-ph is appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a serious paper. It defines NESS and TANESS for infinite Lindblad spin systems in the Jaksic-Pillet manner, proves a new sufficient condition for the thermodynamic and long-time limits to commute, and gives an explicit model showing that spectral gaps alone are not enough. The noncommutation example (1/2 vs 0 for the spin expectation) is exact and convincing.\n\nWhat is actually new: Theorems 8 and 10. If the finite-volume Liouvillians have uniform line (or point) spectral gaps, a uniform exceptional gap, and a uniform bound on the condition number of a Jordanizing map, then for any initial state the infinite-system NESS (or TANESS) exists, is unique, and equals the thermodynamic limit of the finite-volume steady states. The proofs are careful and clearly written. Lemma 11, showing purely imaginary eigenvalues are semisimple, is a nice observation. The model in Section 5 is well chosen: the spectrum is computed exactly, the gaps are uniform, and the exact time evolution gives different results depending on the order of limits.\n\nThe soft spot is Section 5.4. The authors want to show that the example violates assumption (A2)/(B2), i.e., the condition number is unbounded. What they actually prove is that a particular explicitly constructed Jordanizing map has condition number growing like (1-epsilon)^{-2n}. But as they themselves note in Section 4.1, the Jordanizing map is highly non-unique: composing with any invertible map that commutes with the Jordan form changes the condition number. A lower bound for one choice does not rule out another choice with much smaller condition number. So the direct verification that (A2) fails is incomplete. Note, however, that this does not affect the main counterexample: the noncommutation is already proven in Section 5.2 by direct calculation. The paper would be cleaner if Section 5.4 either proved an invariant lower bound over all admissible V, or simply said the constructed map has unbounded condition number and left the connection to (A2) as a remark. This is a moderate, fixable weakness.\n\nThe citation pattern is fine. The reliance on [34] for the thermodynamic limit is standard. The self-citations are relevant, though [44] being 'in preparation' is a little fragile.\n\nWho should read this: anyone working on rigorous open quantum systems or dissipative phase transitions. It is a solid extension of an established program rather than a revolution. I would send it to a serious referee; with a revision of Section 5.4 it would be a good paper.","headline":"Solid operator-algebraic definition and sufficient condition for NESS/TANESS commutation, with an exact noncommutation example; the condition-number claim in §5.4 is under-supported but the counterexample itself stands.","tokens_in":30951,"tokens_out":2629,"would_cite":true,"duration_ms":25821,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82C10","46L55","81S22"],"pacs":[],"model":"deepseek-v4-flash","headline":"A uniform condition-number bound makes thermodynamic and long-time limits commute for infinite open quantum spin systems.","keywords":["Lindblad dynamics","quantum spin systems","nonequilibrium steady state","thermodynamic limit","condition number","spectral gap","Jordan canonical form","C*-algebra"],"falsifier":"Construct a local Lindblad family satisfying Assumption 2 whose finite subsystems have uniform line/point gaps and a uniform exceptional gap, and exhibit Jordanizing maps $V^\\epsilon_\\Lambda$ with uniformly bounded condition numbers; if for some initial state the thermodynamic-then-long-time limit differs from the long-time-then-thermodynamic limit, Theorems 8 and 10 are false. Conversely, in the paper's example one only needs to verify the lower bound $\\kappa_{\\Lambda_n} \\ge (1-\\epsilon)^{-2n}/2$ and the noncommutation (expectation value $1/2$ versus $0$) to confirm that dropping the $\\kappa$","tokens_in":30130,"feed_emoji":"⚛️","tokens_out":6309,"duration_ms":63049,"temperature":0.7,"pith_summary":"The paper asks when the steady state of an infinitely extended open quantum spin system can be computed by first solving finite subsystems and then taking the thermodynamic limit. It defines nonequilibrium steady states (NESS) and time-averaged NESS (TANESS) as cluster points of the infinite-system Lindblad dynamics, and shows the answer is controlled by more than spectral gaps. A sufficient condition is established: if the finite-system Liouvillians have uniform line/point gaps, a uniform exceptional gap for defective eigenvalues, and a uniform upper bound on the condition number of a map putting each normalized Liouvillian into Jordan form, then the thermodynamic and long-time limits commute for every initial state, and the infinite-system NESS/TANESS is unique. The condition number plays the decisive role: the paper exhibits a local model with uniform gaps in which this number grows exponentially with system size and the two limiting orders give different expectation values.","feed_headline":"One number decides when thermodynamic and long-time limits agree","feed_subtitle":"Spectral gaps aren't enough: a uniform nonnormality bound is what makes infinite steady states computable.","key_machinery":"The central object is the condition number $\\kappa_\\Lambda(V^\\epsilon_\\Lambda) = \\|V^\\epsilon_\\Lambda\\|\\,\\|(V^\\epsilon_\\Lambda)^{-1}\\|$ of an invertible linear map $V^\\epsilon_\\Lambda$ that carries the normalized Liouvillian $(\\Delta_{\\rm ex}-\\epsilon)^{-1}L_\\Lambda$ into Jordan canonical form. This number quantifies how far each finite-system Liouvillian is from being normal, i.e., how strongly its defective (non-semisimple) eigenvalues deform the time evolution. Together with three spectral gaps — the line gap $\\Delta_\\Lambda$, the point gap $\\Delta^p_\\Lambda$, and the exceptional gap $\\Delta^{\\rm ex}_\\Lambda$ — the uniform boundedness of $\\kappa_\\Lambda$ controls the prefactor in the fini","core_discovery":"On the paper's own terms: for a family of local Lindblad generators $L_\\Lambda$ acting on finite subsystems of a quasi-local $C^*$-algebra, define the finite-system steady state by projecting onto the zero-eigenvalue spectral subspace $\\Pi_\\Lambda$. The main theorem states that, under the three uniform bounds, for any initial state $\\omega$ the weak-$*$ limit over $\\Lambda$ of $\\omega \\circ \\Pi_\\Lambda$ exists, is independent of the chosen subnet, and equals every NESS and TANESS of the infinite-system dynamics. The proof derives a $\\Lambda$-independent exponential bound $|\\omega \\circ \\gamma^\\Lambda_t(A) - \\omega \\circ \\Pi_\\Lambda(A)| \\le \\kappa\\|A\\| e^{-\\min\\{\\Delta,\\epsilon\\}t}$, obtained","pith_inferences":["Editorial inference: because $V^\\epsilon_\\Lambda$ is non-unique, the theorem is really about the best possible diagonalizing map; a model might satisfy the uniform condition-number assumption via an adapted Jordanization even when a naive basis has exponentially large $\\kappa$, so the practical test is to search for a well-conditioned $V^\\epsilon_\\Lambda$ rather than to compute $\\kappa$ in one fix","Editorial inference: the exponential lower bound $\\kappa_{\\Lambda_n} \\ge (1-\\epsilon)^{-2n}/2$ connects the growth of Jordan-block size to the breakdown of commutativity; one could test whether other driven dissipative chains with non-Hermitian skin-like sensitivity also have size-dependent condition numbers and hence ambiguous infinite-system steady states.","Editorial inference: for closed (unitary) dynamics the Liouvillian is normal, so the condition-number assumption holds automatically; the theorem thus contains, as a special case, the familiar statement that a uniformly gapped unitary dynamics has a unique thermodynamic-limit steady state."],"forward_implications":["If the uniform condition-number bound holds, finite-system NESS/TANESS data (the spectral projection $\\Pi_\\Lambda$) converge weak-$*$ to the unique infinite-system NESS/TANESS, so the infinite-system steady state can be computed algebraically from finite Liouvillians.","The convergence to the steady state is exponential with rate $\\min\\{\\Delta,\\epsilon\\}$, uniformly in system size, making finite-cluster calculations quantitatively reliable whenever the sufficient condition is verified.","For diagonalizable Liouvillians, the line-gap assumption plus a uniform diagonalizer condition number suffices (Corollary 9); the exceptional-gap assumption is then automatic.","When the condition-number bound fails, uniform spectral gaps alone can still guarantee a well-defined infinite dynamics, yet the two orders of limits irreversibly change expectation values, as shown by the paper's example.","The result applies to every initial state, not only stationary or low-energy states, and yields uniqueness of the NESS/TANESS for that initial state."],"supporting_citations":[{"why":"Supplies the Lieb-Robinson bound and the existence of the thermodynamic-limit semigroup $\\gamma^\\Gamma_t$ used throughout the paper.","marker":"[34]"},{"why":"Provides the weak-$*$ compactness of the state space, ensuring NESS and TANESS cluster points exist.","marker":"[9]"},{"why":"Gives the net-convergence formulation used to extend the thermodynamic limit from sequences to nets.","marker":"[10]"},{"why":"Establishes the Lindblad form of the generator for completely positive unit-preserving quantum dynamical semigroups.","marker":"[26]"},{"why":"Provides the Gorini-Kossakowski-Sudarshan form of the Liouvillian for finite-level systems, the basis of the finite-system generators $L_\\Lambda$.","marker":"[13]"},{"why":"Introduces the concept of exceptional points and the stability theory for non-semisimple eigenvalues, underlying the exceptional gap $\\Delta^{\\rm ex}_\\Lambda$.","marker":"[22]"},{"why":"Documents the role of the condition number in controlling spectral stability of nonnormal operators, motivating Assumptions (A2) and (B2).","marker":"[48]"},{"why":"Inspires the proof of Lemma 11 that purely imaginary eigenvalues of a Liouvillian are semisimple and all eigenvalues have non-positive real part.","marker":"[30]"}],"fun_headline_variants":["Nonnormality, not spectral gaps, decides NESS limits","One constant fixes infinite steady states when gaps fail","When gaps aren't enough: the condition number rule","The missing ingredient for infinite Lindblad steady states","A single number gates agreement of thermodynamic and long-time limits"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The load-bearing premise is Assumptions (A2)/(B2): there is a single finite constant $\\kappa$ such that every finite subsystem's Liouvillian can be put into Jordan form by a linear map with condition number at most $\\kappa$; if this uniform bound fails, even perfectly uniform spectral gaps do not guarantee that the thermodynamic and long-time limits commute.","fun_headline_variants_meta":{"raw":{"variants":["Nonnormality, not spectral gaps, decides NESS limits","One constant fixes infinite steady states when gaps fail","When gaps aren't enough: the condition number rule","The missing ingredient for infinite Lindblad steady states","A single number gates agreement of thermodynamic and long-time limits"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000225,"raw_usage":{"total_tokens":1275,"prompt_tokens":691,"completion_tokens":584,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":435,"completion_tokens_details":{"reasoning_tokens":507}},"tokens_in":435,"tokens_out":584,"duration_ms":6850,"temperature":1.0,"reasoning_tokens":507,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T22:09:15.362998+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a local Lindblad family satisfying Assumption 2 whose finite subsystems have uniform line/point gaps and a uniform exceptional gap, and exhibit Jordanizing maps $V^\\epsilon_\\Lambda$ with uniformly bounded condition numbers; if for some initial state the thermodynamic-then-long-time limit differs from the long-time-then-thermodynamic limit, Theorems 8 and 10 are false. Conversely, in the paper's example one only needs to verify the lower bound $\\kappa_{\\Lambda_n} \\ge (1-\\epsilon)^{-2n}/2$ and the noncommutation (expectation value $1/2$ versus $0$) to confirm that dropping the $\\kappa$","supporting_citations":[{"cited_title":"Lieb-Robinson bounds and existence of the thermodynamic limit for a class of irreversible quantum dynamics","cited_arxiv_id":null,"evidence_quote":"Supplies the Lieb-Robinson bound and the existence of the thermodynamic-limit semigroup $\\gamma^\\Gamma_t$ used throughout the paper."},{"cited_title":"Bratteli and D","cited_arxiv_id":null,"evidence_quote":"Provides the weak-$*$ compactness of the state space, ensuring NESS and TANESS cluster points exist."},{"cited_title":"Bratteli and D","cited_arxiv_id":null,"evidence_quote":"Gives the net-convergence formulation used to extend the thermodynamic limit from sequences to nets."},{"cited_title":"On the generators of quantum dynamical semigroups","cited_arxiv_id":null,"evidence_quote":"Establishes the Lindblad form of the generator for completely positive unit-preserving quantum dynamical semigroups."},{"cited_title":"Completely positive dy- namical semigroups of N-level systems","cited_arxiv_id":null,"evidence_quote":"Provides the Gorini-Kossakowski-Sudarshan form of the Liouvillian for finite-level systems, the basis of the finite-system generators $L_\\Lambda$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the concept of exceptional points and the stability theory for non-semisimple eigenvalues, underlying the exceptional gap $\\Delta^{\\rm ex}_\\Lambda$."},{"cited_title":"Trefethen and M","cited_arxiv_id":null,"evidence_quote":"Documents the role of the condition number in controlling spectral stability of nonnormal operators, motivating Assumptions (A2) and (B2)."},{"cited_title":"Spectral theory of Liouvillians for dissipative phase transi- tions","cited_arxiv_id":null,"evidence_quote":"Inspires the proof of Lemma 11 that purely imaginary eigenvalues of a Liouvillian are semisimple and all eigenvalues have non-positive real part."}],"review_version":1}