{"id":"ba75c4a2-acd0-4164-a1de-1e76e10067f8","arxiv_id":"2608.07882","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 driven non-Hermitian SSH chain, a biorthogonal dynamical phase transition occurs across a finite parameter region where the conventional self-normal rate remains analytic, so the two definitions are not equivalent.","lead":"The paper finds a whole parameter region in a driven non-Hermitian chain where one mathematical way of detecting a dynamical phase transition shows sharp signals while another, equally standard way stays smooth. This overturns the earlier belief that the two definitions always switch on together, making the choice of inner product a physically decisive ingredient.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Self-normal existence condition Eq. (14) is asserted without derivation; the finite B-only regime is defined by its complement, so an algebraic error there would directly change the central claim.","rationale":"I agree with the reader's weakest assumption: the derivations of Eqs. (12) and (14) are not shown, and the four-regime diagram rests on them. My independent partial check of the biorthogonal condition confirms that the zero of ⟨~u_-|U_R(t)|u_-⟩ indeed requires Re E = 0 (the condition reduces to |E - ω/2| = |E + ω/2|), so the biorthogonal critical-time formula is on solid ground. The self-normal condition is more delicate: it requires the right eigenstate to have vanishing σ_z expectation, and Eq. (14) is a nontrivial inequality whose derivation is absent. The B-only region is the set where the biorthogonal condition holds but the self-normal condition fails; thus the central claim depends directly on the correctness of Eq. (14). The proposed numerical test would settle whether Eq. (14) correctly marks the self-normal DQPT boundary. Even if minor corrections were needed, the qualitative B-only region would likely survive for η>0, but the exact boundaries and the claimed γ-asymmetry need verification. This supports the reader's CONDITIONAL verdict rather than a rejection.","tokens_in":10868,"tokens_out":28956,"duration_ms":265775,"concrete_test":"On a (η, γ) grid covering Regions I-IV, numerically compute D(k) = ⟨u_{k,-}|σ_z|u_{k,-}\rangle for k ∈ [-π, π] using the exact right eigenstates of H_F(k), and determine whether D(k) has any real zero. Plot the zero-existence boundary and compare with the purple curves in Fig. 4 (Eq. (14)). In particular, verify that at the B-only point (η, γ) = (0.5, 5.5) no k satisfies D(k)=0, while at the Both point (-0.15, 2.6) and S-only point (-1.5, 2.5) at least one k does. Also check that the boundary points satisfy Eq. (14) to within numerical tolerance.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of a finite biorthogonal-only regime is obtained by intersecting the biorthogonal DQPT region (Re E_kc,- = 0, i.e., the exceptional-line-bounded regions) with the complement of the self-normal DQPT region. The self-normal region boundary is fixed by Eq. (14), a piecewise inequality that is stated without derivation. Unlike the biorthogonal critical condition, which can be verified directly from the overlap zero condition (Re B = 0 reduces to Re E = 0), Eq. (14) is a nontrivial statement about the existence of a real solution k to the right-state condition ⟨u_{k,-}|σ_z|u_{k,-}\rangle = 0. The claimed γ → -γ asymmetry makes the branch selection delicate; a missing or mislabeled solution branch in the derivation would shift the B-only / Both / S-only boundaries, and in an extreme case could shrink the B-only area to a non-generic set. Because the overlap zero at t=T/2 requires this condition, and the paper does not show the intervening algebra, the finite extent of the B-only regime rests on an unverified assertion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a periodically driven non-Hermitian Su-Schrieffer-Heeger chain that is exactly solvable in a rotating frame. It defines biorthogonal and self-normal Loschmidt echoes for the same initial state and time evolution, derives analytic critical conditions, and classifies the parameter space into four regimes: biorthogonal-only (B-only), both, self-normal-only, and neither. The central claim is the existence of a finite B-only Floquet DQPT region, in which the biorthogonal rate is nonanalytic while the self-normal rate remains smooth, together with the additional claim that biorthogonal criticality produces a pair of critical times per driving period whereas self-normal criticality produces a single midpoint time.","tokens_in":11012,"tokens_out":40627,"duration_ms":398607,"significance":"If correct, the paper establishes a qualitative distinction between biorthogonal and self-normal DQPTs, showing that the choice of overlap structure can decide whether a dynamical phase transition occurs at all. The manuscript has genuine strengths: the model is exactly solvable, the biorthogonal critical-momentum condition follows transparently from the zero of the overlap and reduces correctly at k=0 and k=pi, the four-regime picture is supported by representative rate-function plots, and no parameter fitting is involved. The B-only regime is a consequence of the derived critical conditions rather than an input. However, two analytic steps that are load-bearing for the headline claims are not shown, and one of them, the paired-critical-time statement, appears to overcount the solutions of the zero-overlap equation.","major_comments":[{"comment":"Solving the biorthogonal zero-overlap condition gives, after factoring out e^{-i omega t/2}, the equation 1 + C e^{i omega t}=0 with C=(E-omega/2)/(E+omega/2). Imposing Re E=0 makes |C|=1, and for a fixed branch of E this equation has exactly one solution per driving period: if Im E>0 then t=tau+nT, while if Im E<0 then t=(n+1)T-tau. It does not produce both tau and T-tau for the same branch. Equation (12) therefore appears to overcount the critical times, and the abstract's 'pair of critical times' claim, as well as the descriptions of Figs. 2 and 3, need to be corrected or supported by a derivation that specifies the branch convention. This is not cosmetic: the paired-time structure is one of the two headline distinctions drawn between the two formulations.","section":"Eq. (12) and Fig. 2"},{"comment":"The piecewise condition for the existence of a real solution k to the self-normal zero-overlap condition is stated without derivation. Because the B-only region in Fig. 4 is the complement of this condition inside the biorthogonally critical part of parameter space, an algebraic or branch-selection error in Eq. (14) would directly change the central claim. Please provide the derivation, and clarify whether the condition uses signed gamma or |gamma|: the eta<=0 branch contains (gamma/omega)^2 while the eta>0 branch contains gamma/omega, and the text's statement that reversing gamma does not preserve the admissible solution needs to be reconciled with this notation.","section":"Eq. (14)"}],"minor_comments":[{"comment":"The title and running header contain artifacts such as 'Transi tions' and other line-break spacing issues; please clean these before submission.","section":"Title and headers"},{"comment":"If gamma is allowed to be negative, the notation gamma/omega in Eq. (14) should be made explicit as signed or absolute; for eta>0 a negative gamma makes the first inequality impossible, which may be intentional but should be stated.","section":"Eq. (14)"},{"comment":"The calculations use N=8000 unit cells, but no finite-size convergence statement is given; since the Loschmidt rates are defined in the thermodynamic limit, a brief note on convergence would be helpful.","section":"Numerics"}],"recommendation":"major_revision","confidential_remarks":"The referee report given to the authors focuses on Eq. (14), but the paired-critical-time issue in Eq. (12) is at least as serious because it appears in the abstract and in the interpretation of the figures. I checked the zero-overlap algebra for a representative k=0 case and found only one solution per period for a fixed branch, not the two listed in Eq. (12). If the authors can justify the pair through a precise branch convention, the paper may be publishable after the missing derivations are supplied; otherwise the temporal-structure claim must be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Honest take: the B-only regime is real and new. The paper shows, in an exactly solvable driven non-Hermitian SSH chain, that the biorthogonal Loschmidt rate can become nonanalytic while the self-normal rate stays smooth for the same parameters, initial state, and evolution. That overturns the working assumption that the two always coexist. The four-regime diagram (B-only, Both, S-only, None) is a useful organizing result, and the analytical connection between biorthogonal critical momenta and real-part gap closings (exceptional lines) is clean. The pair of critical times per period versus a single midpoint time for self-normal is a nice concrete distinction.\n\nThe paper is well executed for a Letter. The model is simple enough that the critical-time formula at k=0 checks out, and the rate-function plots are consistent with the claimed cusps. The derivation of the biorthogonal condition is transparent: the echo vanishes iff the overlap element vanishes, which reduces to a real-part condition. That part is solid.\n\nThe soft spot is real but narrow. The self-normal existence condition, Eq. (14), is stated as a piecewise inequality with no derivation. The B-only regime is defined as the complement of that region intersected with the biorthogonal region. An algebraic error in Eq. (14) would shift the boundaries and could shrink the B-only area. The stress-test note is right that the gamma-to-minus-gamma asymmetry makes branch selection delicate. That said, nothing in the paper suggests an error; the condition is plausible and the plots match the stated boundaries. This is a standard Letter-style compression, but it is a genuine gap: the central claim rests on that inequality. The authors should put the derivation in a supplemental appendix or at least sketch the branch analysis.\n\nFor the audience: anyone working on non-Hermitian DQPTs or Floquet dynamics will want to see this. It is a serious contribution, not a footnote. I'd send it to referees. The missing derivation is exactly the kind of thing a referee should ask for; it is not a reason to desk reject.\n\nRecommended: engage, with a request for the Eq. (14) derivation.","headline":"A clean, exactly solvable demonstration that biorthogonal and self-normal DQPTs need not coexist; the only real soft spot is the under-derived self-normal existence condition, which deserves a referee's look.","tokens_in":11625,"tokens_out":1898,"would_cite":true,"duration_ms":19123,"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 driven non-Hermitian SSH chain, biorthogonal Loschmidt rate can turn nonanalytic while self-normal rate stays analytic.","keywords":["biorthogonal DQPT","Floquet DQPT","non-Hermitian SSH chain","Loschmidt echo","exceptional lines","self-normal Loschmidt rate","dynamical quantum phase transition","associated states"],"falsifier":"Directly compute $g^B_k(t)$ and $g^S_k(t)$ for parameters in the B-only region, for example $(\\eta,\\gamma)=(0.5,5.5)$ with $\\omega=2$, and check whether $\\lambda_B(t)$ shows cusps at the times predicted by Eq. (12) while $\\lambda_S(t)$ remains analytic; equivalently, tune $\\gamma$ across the exceptional line $\\gamma^2=16+\\omega^2$ and verify that biorthogonal cusps appear exactly at the boundary while the self-normal rate stays smooth.","tokens_in":10609,"feed_emoji":"⚛️","tokens_out":5907,"duration_ms":55809,"temperature":0.7,"pith_summary":"This paper tries to establish that, in non-Hermitian quantum systems, the choice of inner product can decide not just where a dynamical quantum phase transition (DQPT) occurs but whether it occurs at all. Concretely, it studies an exactly solvable periodically driven non-Hermitian Su-Schrieffer-Heeger chain and exhibits a finite biorthogonal-only regime: for the same Hamiltonian, initial state, and time evolution, the biorthogonal Loschmidt rate develops nonanalytic cusps while the self-normal Loschmidt rate remains smooth. The critical conditions are derived analytically, tying biorthogonal criticality to real-part gap closings of the effective Floquet Hamiltonian and its exceptional lines, while self-normal criticality has no such spectral boundary. If the claim is right, observed DQPTs in non-Hermitian settings cannot be treated as formulation-independent, and the biorthogonal construction selects genuinely distinct nonequilibrium physics.","feed_headline":"Driven non-Hermitian chain shows biorthogonal-only phase transitions","feed_subtitle":"Finite parameter region has biorthogonal Loschmidt cusps while self-normal rate is smooth.","key_machinery":"The load-bearing object is the normalized biorthogonal Loschmidt echo built from associated (left) states, $g^B_k(t)=|\\langle \\tilde{u}_{k,-}|U_R(t)|u_{k,-}\\rangle|^2 / \\sum_\\mu |\\langle \\tilde{u}_{k,\\mu}|U_R(t)|u_{k,-}\\rangle|^2$, which vanishes only when the associated-state overlap is zero. Working in the rotating frame reduces the driven problem to a time-independent effective Floquet Hamiltonian $H_F(k)=d_k\\cdot\\sigma$, and the zero-overlap condition reduces to $\\mathrm{Re}\\,E_{k_c,-}=0$ with $E_{k_c,-}\\neq 0$, i.e., a real-part gap closing at exceptional lines. This spectral locking is what separates biorthogonal from self-normal criticality.","core_discovery":"The central discovery is a parameter region in a driven non-Hermitian SSH chain where the biorthogonal Floquet DQPT and the self-normal Floquet DQPT decouple completely. Starting from the minus branch of the effective Floquet Hamiltonian, the biorthogonal Loschmidt echo vanishes at momenta where $\\mathrm{Re}\\,E_{k_c,-}=0$ with $E_{k_c,-}\\neq 0$, producing two critical times per driving period; the same initial state gives a self-normal echo whose nonanalyticity requires a different condition and appears only once per period, at the midpoint. The authors solve both conditions analytically and obtain four regimes (B-only, Both, S-only, None), with the B-only region finite rather than fine-tuned. They conclude that the overlap structure is a decisive ingredient in non-Hermitian Floquet dynamical criticality.","pith_inferences":["If the B-only regime survives beyond this exactly solvable model, the biorthogonal Loschmidt rate offers a way to detect exceptional lines dynamically, without directly probing the spectrum.","The $\\gamma\\to-\\gamma$ asymmetry of the self-normal region, contrasted with the symmetric biorthogonal region, suggests the self-normal criterion is branch-dependent; testing initial states occupying the plus branch would be a natural extension.","The photonic-resonator and topolectrical-circuit platforms mentioned in the paper could make the B-only regime testable with current Floquet engineering, with the predicted pair of cusps per period as a clean experimental signature."],"forward_implications":["Biorthogonal and self-normal DQPTs are not necessarily concomitant; a finite B-only regime exists in the $(\\eta,\\gamma)$ plane.","Biorthogonal Floquet DQPTs are tied to real-part gap closings of the effective Floquet Hamiltonian, so exceptional lines mark the onset boundaries.","Each biorthogonal critical momentum yields a pair of critical times per driving period, whereas self-normal criticality yields a single critical time at the midpoint.","The full parameter space is divided into B-only, Both, S-only, and None regimes, with boundaries given by the analytic conditions in Eqs. (12) and (14)."],"supporting_citations":[{"why":"Supplies the biorthogonal DQPT construction, including the associated-state normalized Loschmidt echo that defines $\\lambda_B(t)$.","marker":"[39]"},{"why":"Provides the self-normal Loschmidt echo formulation that $\\lambda_S(t)$ is compared against.","marker":"[40]"},{"why":"Gives the biorthogonal left/right eigenstate framework underlying the associated-state construction.","marker":"[52]"},{"why":"Provides the rotating-frame Floquet solution used to obtain the effective Hamiltonian and the critical-time formulas.","marker":"[59]"},{"why":"Reports experimental self-normal and biorthogonal DQPTs in non-Hermitian quantum walks, motivating the question of whether the two can separate.","marker":"[58]"}],"fun_headline_variants":["Biorthogonal-only Floquet DQPTs: finite regime, two cusps per period","Biorthogonal Loschmidt cusps appear twice per period; self-normal once","Exceptional lines trigger biorthogonal-only Floquet DQPTs","Biorthogonal Floquet DQPTs pin to exceptional lines: self-normal doesn't"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The algebraic derivation of the critical-time formula in Eq. (12) and the self-normal existence condition in Eq. (14) is asserted rather than shown; if either algebraic condition is wrong, the four-regime diagram shifts, even though the qualitative scenario might survive.","fun_headline_variants_meta":{"raw":{"variants":["Biorthogonal-only Floquet DQPTs: finite regime, two cusps per period","Biorthogonal Loschmidt cusps appear twice per period; self-normal once","Exceptional lines trigger biorthogonal-only Floquet DQPTs","Biorthogonal Floquet DQPTs pin to exceptional lines: self-normal doesn't"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001567,"raw_usage":{"total_tokens":6274,"prompt_tokens":979,"completion_tokens":5295,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":595,"completion_tokens_details":{"reasoning_tokens":5196}},"tokens_in":595,"tokens_out":5295,"duration_ms":36651,"temperature":1.0,"reasoning_tokens":5196,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:46:01.050631+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Directly compute $g^B_k(t)$ and $g^S_k(t)$ for parameters in the B-only region, for example $(\\eta,\\gamma)=(0.5,5.5)$ with $\\omega=2$, and check whether $\\lambda_B(t)$ shows cusps at the times predicted by Eq. (12) while $\\lambda_S(t)$ remains analytic; equivalently, tune $\\gamma$ across the exceptional line $\\gamma^2=16+\\omega^2$ and verify that biorthogonal cusps appear exactly at the boundary while the self-normal rate stays smooth.","supporting_citations":[{"cited_title":"Nie, B.-B","cited_arxiv_id":null,"evidence_quote":"Supplies the biorthogonal DQPT construction, including the associated-state normalized Loschmidt echo that defines $\\lambda_B(t)$."},{"cited_title":"Tian, H.-X","cited_arxiv_id":null,"evidence_quote":"Provides the self-normal Loschmidt echo formulation that $\\lambda_S(t)$ is compared against."},{"cited_title":"Fu, Probability balance equation, inner product and Loschmidt echo for non-Hermitian systems, Commun","cited_arxiv_id":null,"evidence_quote":"Gives the biorthogonal left/right eigenstate framework underlying the associated-state construction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports experimental self-normal and biorthogonal DQPTs in non-Hermitian quantum walks, motivating the question of whether the two can separate."}],"review_version":1}