{"id":"a6f58c7a-7c38-4179-92a2-7466faf3ccf0","arxiv_id":"1908.11572","paper_version":3,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The long-time average of the Loschmidt echo (or its rate function) shows nonanalytic behavior when the post-quench parameter crosses a phase transition, and its second derivative equals four times the fidelity susceptibility for small quenches.","lead":"This paper shows that the long-term average of the Loschmidt echo, a measure of how much a quantum state remembers its starting point, develops a sharp kink or a diverging second derivative when a system is suddenly quenched across a phase transition. It offers a simple, time-independent diagnostic for nonequilibrium quantum phase transitions, connected to the well-known fidelity susceptibility.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Sec. II.E small-quench identity is derived for χδ=-∂²Lδ/∂δ², not for the rate-function quantity χλf=-∂²η/∂λf² plotted in Figs. 4-6; the two differ by a factor 1/L and an overall sign, so the abstract's 'connection to fidelity susceptibility' does not justify the central claim.","rationale":"The paper's numerical demonstrations are valuable and likely correct, but the analytical bridge to fidelity susceptibility is not stated for the quantity actually used as the diagnostic. The reader flagged the uncontrolled perturbation theory at criticality; that is a real rigor gap. However, even before reaching the critical point, there is a more direct mismatch: the derivation concerns χδ, the second derivative of the LE average, while the headline diagnostic χλf is the second derivative of the rate function. These differ by a factor of the system size and by a sign, so Eq. 31 cannot, as written, explain the positive peaks in Figs. 4-6. The central numerical claim may survive, and the required correction is a rescaling plus a sign convention, so the overall conditional verdict is unchanged. The condition for acceptance should be broadened to require a corrected derivation that tracks the distinction between Lδ and η and addresses the non-uniformity of the perturbation expansion at the gap-closing point.","tokens_in":10507,"tokens_out":25214,"duration_ms":252703,"concrete_test":"Starting from Eqs. 24-26, compute χλf = -∂²η/∂δ² explicitly with η = -(1/S) log Lδ and compare the result with Eq. 31. Then recompute the Haldane model data in Fig. 7 using χλf rather than χδ; if the corrected quantity does not match 4χF/S with the sign and 1/S scaling from this derivation, the fidelity-susceptibility explanation for the rate-function second derivative plotted in Figs. 4-6 is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II.E derives χδ = -∂²Lδ/∂δ² = 4χF for the long-time averaged Loschmidt echo Lδ (Eqs. 24-27), but the central diagnostic used in Figs. 4-6 and emphasized in the abstract is χλf = -∂²η/∂λf² for the rate function η = -(1/S) log L. These are not interchangeable. Expanding Lδ = 1 - 2Aδ² with A = Σ|Hm0|²/(E0-Em)² gives η ≈ 2Aδ²/S, so χλf ≈ -4χF/S, which has the opposite sign and an 1/S prefactor relative to Eq. 31. Thus even where the first-order perturbation theory is controlled, Eq. 31 establishes a divergence in the LE-average second derivative, not in the rate-function second derivative that the paper plots. The claim that the fidelity-susceptibility connection explains the positive peaks in Figs. 4-6 therefore rests on an unstated and currently unjustified identification of χδ with χλf. This is compounded by the reader's point that the perturbation expansion is not uniform when the gap closes.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the infinite-time average of the Loschmidt echo after a sudden quench, L = Σ_n |⟨ψ_n(λ_f)|Ψ(0)⟩|^4, and its rate function η(λ_f) = -(1/L) log L. The authors present numerical evidence that L or η changes sharply and that χ_{λ_f} = -∂²η/∂λ_f² develops peaks when λ_f crosses a quantum phase boundary in the Aubry-André model, the transverse-field Ising chain, and the Haldane model. They also derive a small-quench identity, χ_δ = 4χ_F, connecting the second derivative of L to the fidelity susceptibility, and use this relation to argue that the rate-function second derivative diverges at criticality.","tokens_in":10799,"tokens_out":7153,"duration_ms":68526,"significance":"If the central claim holds, the long-time-averaged Loschmidt echo would provide a simple, model-independent diagnostic for nonequilibrium quantum phase transitions, and the connection to fidelity susceptibility would give an analytic explanation of the divergence. The paper tests the idea on three different models with known transition points, uses no fitted parameters, and provides an exact integral expression for the Ising rate function in Eq. (15), which are clear strengths. However, the fidelity-susceptibility identity is derived for L, not for the rate function η, so the analytic explanation for the divergences in χ_{λ_f} is currently unsupported.","major_comments":[{"comment":"The identity χ_δ = -∂²L_δ/∂δ² = 4χ_F concerns the long-time-averaged echo L, whereas the quantity plotted in Figs. 4-6 and highlighted in the abstract is χ_{λ_f} = -∂²η/∂λ_f² for the rate function η. Because η = -(1/S) log L, the small-δ expansion L_δ ≈ 1 - 2Aδ² with A = Σ_{m≠0} |H_{m0}|²/(E_0-E_m)² gives η ≈ 2Aδ²/S and hence χ_{λ_f} ≈ -4χ_F/S. This differs from Eq. (31) by an overall sign and a 1/S prefactor, so Eq. (31) does not by itself explain the positive peaks in Figs. 4-6. The authors need to state explicitly how χ_{λ_f} is related to χ_F, or remove the claim that the fidelity-susceptibility connection explains the rate-function divergence.","section":"Sec. II.E, Eq. (31) vs Eq. (23)"},{"comment":"The perturbation expansion in δ assumes non-degenerate eigenstates with finite energy denominators. At a quantum critical point the gap closes, so the first-order expansion is not controlled for the modes that dominate the fidelity susceptibility. Using Eq. (31) to argue for a divergence of χ at criticality therefore interchanges the limits δ→0 and the approach to the critical point. The authors should provide a non-perturbative argument or explicitly restrict the identity to quenches away from criticality.","section":"Sec. II.E, Eqs. (25)-(27)"},{"comment":"The claim that χ_{λ_f} diverges in the thermodynamic limit rests on Fig. 6, which shows peak heights increasing with lattice size. The paper does not provide a scaling collapse, an extrapolation to L→∞, or a comparison with the 1/S prefactor that arises from the corrected rate-function relation. The finite-size trends are suggestive but do not by themselves establish a true divergence, especially because the small-quench analysis indicates that χ_{λ_f} and χ_δ scale differently with system size.","section":"Sec. III, Fig. 6"}],"minor_comments":[{"comment":"Equations (27) and (30) omit the summation over m≠0 that appears in Eq. (26); as written, they read as if only a single matrix element contributes. The sums should be restored.","section":"Eq. (27) and Eq. (30)"},{"comment":"The symbol L is used both for the lattice size and for the long-time-averaged Loschmidt echo, which makes expressions such as η = -(1/L) log L confusing. A different symbol, such as S or N, for the system size would improve readability.","section":"Eq. (5) and throughout"},{"comment":"The sign convention for χ_{λ_f} should be checked against the plotted quantity. The text defines χ_{λ_f} = -∂²η/∂λ_f², but the figures appear to show positive peaks; if the plotted quantity is actually ∂²η/∂λ_f² or |χ_{λ_f}|, this should be stated in the captions.","section":"Eq. (23), Figs. 4-6"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a serious referee. The paper proposes using the long-time average of the Loschmidt echo—an inverse participation ratio of the initial state in the eigenbasis of the post-quench Hamiltonian—as a time-independent diagnostic for nonequilibrium quantum phase transitions. It demonstrates the idea in three models (AA, Ising, Haldane) and shows that the rate function η or its second derivative χλf develops a sharp feature when the post-quench parameter crosses a known phase boundary. The numerics are clean: no fitted parameters, benchmarks against known transition points, and a finite-size peak growth in Fig. 6. The neat analytical result is Eq. (31), χδ = 4χF, which is derived and verified numerically in Fig. 7.\n\nThe problem is that Eq. (31) is about χδ = -∂²Lδ/∂δ², while the divergent quantity in Figs. 4–6 is χλf = -∂²η/∂λf², where η = -(1/L) log L. These are not interchangeable. In the small-quench limit, Lδ ≈ 1 - 2Aδ², so η ≈ 2Aδ²/L and therefore χλf ≈ -4χF/L. That has the opposite sign and a 1/L factor relative to χδ. The paper never acknowledges this mismatch, and the abstract's promise about 'the connection between the second derivative of rate function and fidelity susceptibility' is therefore stronger than what is actually shown. The perturbation expansion behind Eq. (31) also loses control at the critical point, though it is fine for finite systems with a small δ. These issues are addressable: either derive the analogous statement for χλf, or explicitly restrict the fidelity-susceptibility relation to χδ and adjust the framing.\n\nMinor points: Eq. (15) has a typo in the integrand, and the finite-size analysis is qualitative, without error bars or scaling collapse. None of this changes the empirical observation that the long-time averaged LE marks the phase boundaries. The paper is useful for people working on quench dynamics and quantum criticality; it shows a simple observable that works across different models. I would send it to peer review and ask for a revision that fixes the χδ/χλf confusion before publication.","headline":"Solid numerical diagnostics for NQPT via long-time averaged LE, but the fidelity-susceptibility connection is derived for the wrong object (χδ vs χλf) and should be reframed.","tokens_in":11296,"tokens_out":6073,"would_cite":false,"duration_ms":54896,"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":"The long-time average of the Loschmidt echo is a detector of nonequilibrium quantum phase transitions.","keywords":["Loschmidt echo","long-time average","nonequilibrium quantum phase transition","quantum quench","fidelity susceptibility","rate function","dynamical quantum phase transition","nonanalyticity"],"falsifier":"Take the transverse-field Ising model with initial field $h_i=0$, evaluate the analytic rate function $\\eta(h_f)$ on progressively larger chains, and compute the second derivative $\\chi_{h_f}$ near $h_f/J=1$; if the peak height does not grow without bound as $L\\to\\infty$, the claimed critical divergence is absent. A complementary check is to compute $\\chi_\\delta$ and $4\\chi_F$ exactly for small but finite $\\delta$ at the critical point and see whether the proportionality $4$ survives when the perturbative expansion fails.","tokens_in":10332,"feed_emoji":"⚛️","tokens_out":9253,"duration_ms":77735,"temperature":0.7,"pith_summary":"This paper establishes that the long-time average of the Loschmidt echo—the time-averaged return probability of a state after a sudden quantum quench—carries a sharp signature of quantum phase transitions in the post-quench Hamiltonian. Across three representative models (Aubry-André, transverse-field Ising, and Haldane), the echo average or its rate function develops a nonanalytic feature when the final quench parameter crosses a critical point, independent of which phase the system starts in. For quenches of small amplitude, the paper derives a proportionality between the second derivative of the echo average and the fidelity susceptibility, tying the dynamical signal to a well-established ground-state probe. The payoff is a time-independent, computable observable that can reveal nonequilibrium quantum criticality even when the usual dynamical quantum phase transitions—zeros of the echo at critical times—are absent or masked.","feed_headline":"Long-time Loschmidt echo pinpoints quantum phase transitions","feed_subtitle":"Across three models, the time-averaged echo or its second derivative turns nonanalytic exactly at the critical point.","key_machinery":"The load-bearing object is the long-time average of the Loschmidt amplitude, $\\overline{\\mathcal{L}}(\\lambda_f)=\\sum_n |\\langle \\psi_n(\\lambda_f)|\\Psi(0)\\rangle|^4$, which has the form of an inverse participation ratio of the initial state in the post-quench eigenbasis; its intensive logarithm is the rate function $\\eta$. The bridge to known quantum-critical physics is the small-quench perturbation theory: expanding the eigenstates of $H(\\lambda+\\delta)$ to first order in $\\delta$ gives $\\overline{\\mathcal{L}}_\\delta = 1 - 2\\delta^2 \\sum_{m\\neq 0} |H_{m0}|^2/(E_0-E_m)^2$, while the ground-state fidelity susceptibility is $\\chi_F = \\sum_{m\\neq 0}|H_{m0}|^2/(E_0-E_m)^2$, yielding $\\chi_\\delta=4\\chi_F$. This identity is what converts a time-averaged dynamical quantity into a ground-state criticality detector.","core_discovery":"The central claim is that the long-time average $\\overline{\\mathcal{L}}(\\lambda_f)$ of the Loschmidt echo, or its rate function $\\eta(\\lambda_f)=-(1/L)\\log \\overline{\\mathcal{L}}(\\lambda_f)$, is a nonequilibrium order parameter in quench-parameter space: it is nonanalytic exactly when $\\lambda_f$ crosses a quantum phase-transition point. The nonanalyticity appears as a sharp change in $\\overline{\\mathcal{L}}$ or $\\eta$ in the Aubry-André and Ising models, and, in the Haldane model where $\\eta$ looks smooth, as a diverging peak in the second derivative $\\chi_{\\lambda_f}=-\\partial^2\\eta/\\partial\\lambda_f^2$. In the small-quench limit $\\lambda_f=\\lambda_i+\\delta$, the paper derives $\\chi_\\delta=-\\partial^2\\overline{\\mathcal{L}}_\\delta/\\partial\\delta^2 = 4\\chi_F$, with $\\chi_F$ the fidelity susceptibility, so the divergence at criticality follows from the known divergence of $\\chi_F$. The signature is shown to be independent of the initial phase, and finite-size data show the peak height growing with system size, consistent with a true thermodynamic-limit divergence.","pith_inferences":["A natural extension is to use the measured long-time averaged return probability for small quenches as an experimental estimator of fidelity susceptibility, which is normally inferred from ground-state wavefunction overlaps.","The same logic may apply to other time-averaged overlaps, such as survival probabilities of excited states or spin autocorrelators, giving a family of nonequilibrium criticality detectors.","For finite quench amplitude $\\delta$, corrections to $\\chi_\\delta=4\\chi_F$ should appear; tracking how the peak position and scaling exponent change with $\\delta$ would test how robust the fidelity-susceptibility link is away from the perturbative regime.","Because the long-time average removes time dependence, the signature may be measurable in platforms with limited coherence times, as long as the time average converges before decoherence sets in; this is an experimentally testable prediction."],"forward_implications":["The long-time averaged Loschmidt echo gives a time-independent probe of nonequilibrium quantum phase transitions, complementing dynamical quantum phase transitions that require zeros of the echo at special times.","The second derivative $\\chi_{\\lambda_f}$ can expose topological phase transitions in models such as the Haldane model, where the rate function itself appears analytic.","Because the signature is independent of the initial phase, it works for quenches starting on either side of the critical point.","The proportionality $\\chi_\\delta=4\\chi_F$ means the divergence of the fidelity susceptibility at criticality is inherited by the quench-averaged return probability."],"supporting_citations":[{"why":"Earlier study of dynamical quantum phase transitions in the Aubry-André model; its DQPT-zero signature is contrasted with the long-time-average signature studied here.","marker":"[33]"},{"why":"Defines dynamical quantum phase transitions through zeros of the Loschmidt echo and nonanalytic dynamical free energy, the phenomenon this paper complements.","marker":"[22]"},{"why":"Establishes the divergence of fidelity susceptibility at quantum phase transitions, the property inherited by the echo's second derivative through the derived proportionality.","marker":"[9]"},{"why":"Shows that ground-state fidelity drops sharply at quantum critical points, providing the conceptual link from wavefunction overlap to phase-transition detection.","marker":"[7]"},{"why":"Identifies inverse participation ratios as measures of Hilbert-space localization, giving the long-time echo average its interpretation as an inverse participation ratio.","marker":"[46]"},{"why":"Defines the Haldane model and its Chern-number phase diagram, whose phase boundaries are the locations of the divergent peaks in the second derivative.","marker":"[48]"},{"why":"Connects dynamical quantum phase transitions to steady-state transitions, supplying the nonequilibrium context in which the paper places the long-time-average signature.","marker":"[43]"}],"fun_headline_variants":["Long-time echo average detects quantum critical points","Time-averaged Loschmidt echo reveals phase transitions","Nonanalytic echo average marks quantum phase boundaries","Second derivative of echo rate signals critical divergence","Quantum phase transitions from long-time echo average"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The small-quench identity $\\chi_\\delta=4\\chi_F$ rests on first-order perturbation theory in $\\delta$, which requires nonzero energy gaps between the ground state and all excited states; at the critical point the gap closes, so the expansion is not controlled for the modes that would produce the divergence.","fun_headline_variants_meta":{"raw":{"variants":["Long-time echo average detects quantum critical points","Time-averaged Loschmidt echo reveals phase transitions","Nonanalytic echo average marks quantum phase boundaries","Second derivative of echo rate signals critical divergence","Quantum phase transitions from long-time echo average"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000221,"raw_usage":{"total_tokens":1444,"prompt_tokens":930,"completion_tokens":514,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":546,"completion_tokens_details":{"reasoning_tokens":455}},"tokens_in":546,"tokens_out":514,"duration_ms":5436,"temperature":1.0,"reasoning_tokens":455,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:11:48.023231+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the transverse-field Ising model with initial field $h_i=0$, evaluate the analytic rate function $\\eta(h_f)$ on progressively larger chains, and compute the second derivative $\\chi_{h_f}$ near $h_f/J=1$; if the peak height does not grow without bound as $L\\to\\infty$, the claimed critical divergence is absent. A complementary check is to compute $\\chi_\\delta$ and $4\\chi_F$ exactly for small but finite $\\delta$ at the critical point and see whether the proportionality $4$ survives when the perturbative expansion fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier study of dynamical quantum phase transitions in the Aubry-André model; its DQPT-zero signature is contrasted with the long-time-average signature studied here."},{"cited_title":"Calabrese, F","cited_arxiv_id":null,"evidence_quote":"Defines dynamical quantum phase transitions through zeros of the Loschmidt echo and nonanalytic dynamical free energy, the phenomenon this paper complements."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the divergence of fidelity susceptibility at quantum phase transitions, the property inherited by the echo's second derivative through the derived proportionality."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that ground-state fidelity drops sharply at quantum critical points, providing the conceptual link from wavefunction overlap to phase-transition detection."},{"cited_title":"Wang and Gao Xianlong Phys","cited_arxiv_id":null,"evidence_quote":"Identifies inverse participation ratios as measures of Hilbert-space localization, giving the long-time echo average its interpretation as an inverse participation ratio."},{"cited_title":"Deluca and A","cited_arxiv_id":null,"evidence_quote":"Defines the Haldane model and its Chern-number phase diagram, whose phase boundaries are the locations of the divergent peaks in the second derivative."},{"cited_title":"Many-Body Localization Transition, Temporal Fluctuations of the Loschmidt Echo, and Scrambling","cited_arxiv_id":"1702.00445","evidence_quote":"Connects dynamical quantum phase transitions to steady-state transitions, supplying the nonequilibrium context in which the paper places the long-time-average signature."}],"review_version":1}