{"id":"acf21117-2384-45e9-825a-f497c77ed28c","arxiv_id":"2602.00230","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Random quantum resets turn Kibble-Zurek scaling into anti-Kibble-Zurek growth, with optimal annealing time scaling as r^{-2/3}.","lead":"This paper asks what happens when a quantum system being slowly driven across a phase transition is randomly reset to its starting state. It finds that resets flip the usual slowing-down picture: defects stop decreasing with quench time and instead grow, with the same universal scaling laws seen for random noise.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (8) is the renewal average only if a reset restarts the ramp; the literal protocol in Eq. (7) resets only the state. The exact average for state-only resets contains U_k(t,s)ρ_k(0)U_k†(t,s), which is not ρ0,k(t−s) for a time-dependent drive. All exponents inherit this.","rationale":"The paper's r=0 limit and the exact TFI-chain solution are solid, and the qualitative crossover mechanism is plausible. My concern is more basic than finite-size fitting: Eq. (8) is introduced as a result of averaging Eq. (7), but the displayed convolution is the renewal average only when a reset at s leaves subsequent evolution equivalent to evolution from t=0. For a linear ramp with continuing drive, the unitary propagator from s to t is not U(t-s,0); hence Eq. (8) does not follow from Eq. (7). If the authors intended QR to include rewinding the drive to h_i, then Eq. (7) is under-specified and the physical claim that QR models an environment is weakened, since a real environment does not rewind the control field. Because Eq. (8) feeds directly into Figs. 1-3 and Eq. (9), every reported exponent and the claimed identity with Gaussian-noise ramps depends on this step. A direct simulation of the literal protocol would settle the issue. The reader's finite-size/fitting concern is legitimate but secondary; even perfect N→∞ data from the wrong average would not support the conclusion. The omitted derivation of Eq. (8) from Eq. (7) (only citation to [25]) is where the gap sits, and the data-availability statement prevents independent numerical verification.","tokens_in":8822,"tokens_out":14103,"duration_ms":155378,"concrete_test":"Implement Eq. (7) literally (state-only resets, ramp continuing) for N=1000 with the paper's h_i=2, h_f=0 and r=2×10^-4, at τ=10, 10^2, 10^3. Use quantum-jump Monte Carlo with ≥10^4 trajectories per point, or solve exactly the renewal formula with U_k(t,s) given above mode by mode. Compare n_r with the Eq. (8)-based values in Fig. 1. If they disagree beyond statistical error, Eq. (8) is invalid and the exponents do not follow. If the intended protocol was drive-restarting resets, amend Eq. (7) accordingly and rerun; then Eq. (8) is justified and the numerical exponents need independent finite-size checks.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Take the manuscript's literal protocol: in Eq. (7), with probability r dt the mode is set to |ψ_k(0)⟩, otherwise propagated by H_k(t). H_k(t) = 2(h(t)-cos k)σ_z + 2 sin k σ_x is explicitly time-dependent and is not reset. For Poissonian resets, conditioning on the last reset at time s gives ρ_r,k(t) = e^{-rt} ρ0,k(t) + r∫_0^t e^{-r(t-s)} U_k(t,s) ρ_k(0) U_k†(t,s) ds, with U_k(t,s)=T exp(-i∫_s^t H_k(u)du). Equation (8) instead has ρ0,k(s) inside the convolution. These agree only if U_k(t,s)=U_k(t-s,0), i.e. if H_k(u+s)=H_k(u) — false for a linear ramp — or if a reset rewinds the drive to h_i. The text's phrase \"resets of the driving\" suggests the latter, but Eq. (7) does not implement it. Since Eq. (8) generates every n_r curve in Fig. 1, the reported α≈1, α′≈3, γ≈2/3, the data collapse, and the claimed equivalence to Gaussian-noise ramps all rest on an average over the wrong stochastic process unless the drive is restarted. This is a correctness gap in the central derivation, not merely a robustness issue.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a linear quench across the quantum critical point of the transverse-field Ising chain subject to Poissonian quantum resets. It claims that resets induce a crossover from Kibble-Zurek scaling (n_0 ~ τ^{-1/2}) to anti-Kibble-Zurek scaling, with defect densities described by Eq. (9), optimal annealing times τ_opt,r ~ r^{-γ} with γ≈2/3, and exponents identical to those found for noisy ramps. The claims are based on exact fermionization and numerical evaluation of a renewal master equation, Eq. (8), for a chain with N=1000 sites.","tokens_in":9233,"tokens_out":9734,"duration_ms":114823,"significance":"If the claimed universality holds, the equivalence between quantum-reset dynamics and uncorrelated/colored noise would be a notable addition to out-of-equilibrium universality. The paper exploits the exact solvability of the TFI chain, and the data collapse in Fig. 1(c) is visually suggestive. However, the central master equation is not consistent with the stated reset protocol, and the universal exponents are inferred from fits to the same numerical data they are then used to support. The significance is therefore conditional on resolving these issues.","major_comments":[{"comment":"Equation (8) does not follow from the protocol defined by Eq. (7). For state-only resets, the state after the last reset at time s is U_k(t,s)|ψ_k(0)⟩, so the renewal average contains r∫_0^t e^{-r(t-s)} U_k(t,s)ρ_k(0)U_k†(t,s) ds. Eq. (8) reduces to this only if U_k(t,s)=U_k(t-s,0), i.e. only if a reset also restarts the linear ramp. The text's phrase 'resets of the driving' may intend this, but Eq. (7) does not implement it. Since all numerical results are obtained from Eq. (8), the reported exponents α, α′, β, γ and the claimed equivalence to noisy ramps describe a different stochastic process from the one defined. Please correct the protocol statement; if state-only resets are intended, the numerical exponents must be recomputed using the correct renewal average.","section":"Topological defects from quenching with QR, Eqs. (7)–(8)"},{"comment":"The central exponents are obtained by fitting the same data they are then used to support. There is no finite-size analysis (N=1000 only), no specified fit ranges, no error bars, and the amplitude h(r) and crossover function Λ(τ) in Eq. (9) are never defined. The collapse in Fig. 1(c) relies on the fitted shift δ_r = ln(r×10^5), whose reference scale is arbitrary. Thus the claimed universal scaling laws are not independently verified. Please provide finite-size scaling, explicit fit windows, and either an analytic derivation or an independent check of the values α≈1, α′=3.000, β≈1/2, and γ≈2/3.","section":"Results, Eq. (9) and Figs. 1–2"},{"comment":"The piecewise form in Eq. (9), with a sharp exponent jump from α′≈3 to α≈1 at τ≈1, is not justified in the manuscript. The text describes this as 'suggestive of some type of criticality,' but no mechanism is given and Λ(τ) is never specified. This makes the two-branch law difficult to falsify. At minimum, the authors should define Λ(τ), specify how the τ<1 and τ>1 regions are chosen, and show a collapse of the full n_r, not only of δn_r after a fitted shift.","section":"Eq. (9) and the τ≈1 crossover"}],"minor_comments":[{"comment":"Please define t′ consistently. If t′ is the time elapsed since the last reset, the survival factor should be written as e^{-r t′} with the substitution u=t−s stated; if t′ is the absolute reset time, a factor e^{rt′} is missing. The current notation is confusing.","section":"Eq. (8)"},{"comment":"The relation δ_r = ln(r×10^5) depends on an arbitrary reference scale. State the units of r and clarify whether δ_r is taken as a fit parameter or derived from the model.","section":"Fig. 1(b) inset"},{"comment":"The coefficient b is identified with c^{-1/2}, where c is the quasiparticle speed. This relation should be derived or explicitly referenced; it appears without derivation.","section":"Eq. (9)"},{"comment":"The statement that n_0 ∼ τ^{-1/2} for τ≳1 is not accompanied by a fit range or accuracy estimate. Given that the paper’s claims rest on numerical scaling, providing these details would improve reproducibility.","section":"Fig. 1(a)"},{"comment":"Since the central results are numerical and obtained from fits, making the data and/or code available would substantially strengthen the paper.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The main technical issue is the inconsistency between the reset protocol in Eq. (7) and the renewal average in Eq. (8). If the authors intended to reset the driving field as well as the state, a clearly revised protocol statement may suffice. If state-only resets are claimed, the numerical results need to be recomputed. The fit-based determination of the exponents also needs more scrutiny before acceptance; I would ask the editor to require finite-size scaling and explicit analysis details."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth one careful read, but the first thing to check is the derivation of Eq. (8), not the fits. The claimed new result—quantum resetting turns Kibble-Zurek scaling into anti-Kibble-Zurek scaling, with optimal annealing times scaling as r^{-2/3} and the same exponents as Gaussian noise—is genuinely new and, if true, would connect resetting protocols to the noise class in a clean way. The TFI setup is standard, the r=0 limit gives KZ correctly, and the qualitative picture of a competition between adiabaticity and resets is plausible.\n\nThe problem is that Eq. (8) is not the average over the protocol in Eq. (7). Eq. (7) resets the mode's state, not the time-dependent Hamiltonian. For a linear ramp the exact renewal average must contain U_k(t,s) ρ_k(0) U_k^†(t,s), and that is not ρ0,k(t-s) unless H_k is time-independent or the reset also rewinds the ramp. The text says 'resets of the driving' at one point, which might mean the latter, but Eq. (7) does not implement it, and Eq. (6) at t_f=2τ with h_f=0 would have to be rethought if the ramp restarts stochastically. So as written, every number in Fig. 1 and Fig. 2 is computed from a different model than the one defined.\n\nThere are also the usual secondary concerns: the exponents are extracted by fitting the same data they are meant to support, there are no error bars from finite-size analysis, only N=1000, and no code or data are released. The τ<1 exponent α'=3.000 is fitted to a signal the authors themselves call experimentally invisible, which is a weak basis for a 'universal' claim.\n\nIf the intended protocol is restart-the-drive resetting, then Eq. (8) may be defensible and the qualitative story is interesting. But as it stands, the text does not say that, and the central derivation is broken. This is not beyond repair, but it needs a corrected derivation, a clear statement of the reset protocol, and ideally an analytic argument for at least γ=2/3.\n\nI would send this to a serious referee, not desk-reject it: the question is timely and the flaw is identifiable and fixable. I would not cite the exponents yet.","headline":"A potentially interesting anti-KZ mechanism, but the central equation doesn't match the stated reset protocol; the exponent claims are unsupported as written.","tokens_in":9813,"tokens_out":7537,"would_cite":false,"duration_ms":92473,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Random quantum resets during a quench convert Kibble-Zurek defect scaling into anti-Kibble-Zurek scaling, with universal exponents identical to those produced by Gaussian noise.","keywords":["quantum resetting","Kibble-Zurek mechanism","anti-Kibble-Zurek scaling","transverse-field Ising chain","topological defects","quantum annealing","universal exponents","out-of-equilibrium dynamics"],"falsifier":"Repeat the same reset-plus-ramp simulation with reset waiting times drawn from a power-law distribution (non-Poissonian) and check whether the optimal-time exponent γ remains 2/3; alternatively, recompute the fitted exponents α, β, γ at N=4000 and N=8000 to test whether they are stable under finite-size scaling.","tokens_in":8635,"feed_emoji":"⚛️","tokens_out":9043,"duration_ms":84247,"temperature":0.7,"pith_summary":"This paper asks whether random interruptions of a quantum quench—reset events that return the system to its initial state—change the universal way in which defects are produced when a system is driven across a quantum critical point. Using the exactly solvable transverse-field Ising chain, the authors find that quantum resetting at rate r shifts the defect density from Kibble-Zurek scaling (defects decrease with slower ramps) to anti-Kibble-Zurek scaling (defects increase with slower ramps), with a crossover at an optimal annealing time. The optimal time and the minimal defect density obey power laws in r that match, within error bars, the exponents found for quenches with uncorrelated or fast colored noise. The paper therefore establishes that Poissonian resets and Gaussian noise produce the same defect statistics, a correspondence with direct consequences for interpreting quantum annealing experiments.","feed_headline":"Resets flip defect scaling to anti-Kibble-Zurek, matching noise","feed_subtitle":"Defect density grows as τ for slow quenches; optimal anneal time scales as reset rate to -2/3.","key_machinery":"The engine of the calculation is the reset-averaged density matrix ρ_r,k(t) = r ∫_0^t e^{-rt'} ρ_{0,k}(t') dt' + e^{-rt} ρ_{0,k}(t), which weights the no-reset evolution by e^{-rt} and each reset event by the Poissonian rate r. Inserting this into the defect-counting formula reduces the problem to the exactly solvable Landau-Zener dynamics of each momentum mode in the fermionized Ising chain. The competition between quench-driven excitations (suppressed as τ grows) and reset-driven excitations (growing with τ) is what produces the crossover and the optimal annealing time.","core_discovery":"The central claim is that the defect density n_r after a linear quench with quantum resetting obeys n_r ≈ h(r) τ^α + b τ^{-β} for slow quenches (τ > 1), with α ≈ 1 and β ≈ 1/2, so the reset-driven part grows linearly with the quench time scale while the ordinary Kibble-Zurek part decays as τ^{-1/2}; their competition produces a local minimum at an optimal time τ_opt,r ∼ r^{-2/3}, with the minimum density scaling as r^{1/3}. The same universal exponents appear in the mean excess energy, and the scaling curves for different reset rates collapse onto a single master curve. This is exactly the behavior previously obtained for quenches subject to Gaussian white or fast colored noise, which leads","pith_inferences":["The paper reports a sharp change of the QR-induced exponent from 3 to 1 at τ ≈ 1 but does not analyze its origin; this may signal a genuine dynamical transition in the reset process and is a natural target for a dedicated analytic or finite-size study.","A testable extension beyond the Poissonian assumption: any reset process with finite mean waiting time may fall into the same universality class, whereas heavy-tailed reset distributions could produce different exponents—this is not claimed by the paper but is consistent with its logic.","The equivalence with Gaussian noise suggests that random resets may be the minimal effective description of a broad class of environmental disturbances during quenches; a numerical experiment with non-Markovian or state-dependent resets could probe how far the universality extends."],"forward_implications":["In a quantum annealer where the environment acts through effective resets, the optimal anneal time grows as r^{-2/3} with the reset rate, so faster resets require faster quenches to minimize defects.","The minimal achievable defect density scales as r^{1/3}, so reducing the reset rate pays off with a power-law reduction in residual defects.","The data collapse means that measuring the defect density at a few reset rates determines the entire universal curve for all rates.","If the paper is correct, tests of Kibble-Zurek scaling in open quantum systems must treat reset-like decoherence and additive noise on the same footing, since both give the same anti-Kibble-Zurek signature in defect density and excess energy."],"fun_headline_variants":["Quantum resets flip defect scaling to anti-Kibble-Zurek","Reset rate sets optimal anneal time for defect minimum","Random resets match noise in quench defect scaling","Anti-Kibble-Zurek from quantum resetting during quench"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The paper assumes that the environment acts as ideal, instantaneous, Poissonian resets of the entire chain to its initial state, and that the averaged density matrix is exactly the weighted integral in Eq. (8), with exponents extracted from N=1000 numerics and no finite-size scaling analysis.","fun_headline_variants_meta":{"raw":{"variants":["Quantum resets flip defect scaling to anti-Kibble-Zurek","Reset rate sets optimal anneal time for defect minimum","Random resets match noise in quench defect scaling","Anti-Kibble-Zurek from quantum resetting during quench"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000185,"raw_usage":{"total_tokens":1148,"prompt_tokens":726,"completion_tokens":422,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":470,"completion_tokens_details":{"reasoning_tokens":360}},"tokens_in":470,"tokens_out":422,"duration_ms":4906,"temperature":1.0,"reasoning_tokens":360,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T06:07:03.214958+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the same reset-plus-ramp simulation with reset waiting times drawn from a power-law distribution (non-Poissonian) and check whether the optimal-time exponent γ remains 2/3; alternatively, recompute the fitted exponents α, β, γ at N=4000 and N=8000 to test whether they are stable under finite-size scaling.","supporting_citations":[],"review_version":1}