{"id":"f4c6dbd9-e477-4239-a8e8-aab8d486326c","arxiv_id":"2412.05656","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Robust control pulses for interacting spin chains are computed in a polynomial-sized operator space, maintaining about 99.9 percent fidelity under up to 5 percent static coupling errors.","lead":"This paper designs control pulses for spin chains that stay accurate even when the coupling strengths are off by up to 5 percent. It does this without simulating the full quantum state, which lets it handle chains of 30 or more spins, a step toward noise-resilient quantum hardware.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claim of robust parasitic control for up to 20 spins is unsupported: exact verification stops at n≤10, and for n>10 the paper infers robustness from the constraint ratio C1/C0, which is not an absolute magnitude and does not bound the first-order error.","rationale":"The reader's verdict is CONDITIONAL, and the most load-bearing concern I identify is the unsupported extrapolation of parasitic robustness from n≤10 to n=20. The paper is transparent about the limitation (Section V.B explicitly says exact verification is limited to n≤10), but then proceeds to make a central claim about up to 20 spins. The inference from C1/C0 is mathematically incomplete: the constraint C of Eq. (27) is a first-order measure, and a ratio of two C values says nothing about the absolute size of the parasitic effect. The unperturbed infidelity shown for n>10 is irrelevant to the parasitic robustness. This is a concrete, addressable gap that does not invalidate the n≤10 results or the general IQC framework, but it does mean the headline scaling claim is not yet established. The finite-sample robustness concern identified by the reader is legitimate but less decisive for the paper's unique contribution; it applies to the average-fidelity claims and could be mitigated by worst-case search, whereas the parasitic n>10 claim currently has no exact or approximate fidelity estimate at all. I therefore recommend keeping the CONDITIONAL verdict, with an explicit condition that the authors either provide exact verification for n>10 (feasible with sparse Lanczos up to n=20) or soften the claim to n≤10. Agreement with the reader is partial because their stated weakest assumption was the sampling issue, although their rationale does mention the parasitic extrapolation.","tokens_in":15182,"tokens_out":7614,"duration_ms":72898,"concrete_test":"Using the optimized pulses from the repository (Ref. [54]) for cluster-state preparation with parasitic ZZ correction, compute for n=12, 14, 16, 18, and 20 the exact average infidelity of Eq. (3) via Lanczos propagation over 1000 random parasitic error vectors drawn uniformly from [-0.05g, 0.05g]^(n-1). If the average infidelity exceeds 1e-3 for any n>10, the 'up to 20 spins' claim fails. Also compare these computed infidelities against the reported C1/C0 values to test whether the ratio is a reliable proxy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section V.B verifies robustness against parasitic ZZ interactions exactly only for n≤10 using Lanczos propagation in the full Hilbert space. For n>10, the paper states that 'robustness can be inferred from the magnitude of the constraint C as it is a measure of the perturbation’s effect,' but Figure 5 reports only the ratio C1/C0 and the unperturbed infidelity, not the absolute value of C1 nor the parasitic infidelity. A small ratio does not imply a small first-order error unless C0 is itself small; the paper provides no calibration connecting C1 values to actual infidelities for n>10. The abstract's central claim of 'explicit control solutions for spin chains with parasitic interactions of up to 20 spins' therefore rests on an extrapolation that is not substantiated. The finite-sample robustness issue (60 training/1000 validation points) is a secondary concern for the 99.9% fidelity headline, but the n>10 parasitic gap is the more immediate missing link.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript extends the implicit quantum control (IQC) framework of Ref. [46] to robust pulse design for spin chains. The system Hamiltonian in Eq. (6), with nearest-neighbor XX couplings, local Z fields, and end-chain drives, is used to prepare cluster states, GHZ states, and to implement a collective-to-single-qubit measurement mapping. Robustness against static fluctuations of the coupling constants is obtained by optimizing an ensemble-averaged operator infidelity (Eqs. (3) and (8)). For parasitic ZZ interactions, the paper introduces a first-order perturbative constraint C (Eq. (27)) that is jointly minimized with the noiseless infidelity. Exact verification via Lanczos propagation is presented for n≤10, and the authors infer robustness for chains up to n=20 from the ratio C1/C0 and the unperturbed infidelity in Fig. 5.","tokens_in":15378,"tokens_out":6137,"duration_ms":60987,"significance":"If the scaling claims are correct, the paper makes a useful step: it indicates that robust optimal control for interacting spin chains of tens of qubits can be approached with quadratic numerical effort, and it provides code and optimized pulses in a public repository. The exact verification for n≤10 and the explicit perturbative construction in Sec. V.A are concrete strengths. However, the headline claims go beyond what is directly verified: the n>10 parasitic result is an extrapolation from a constraint ratio, and the 99.9% robustness statement rests on averages over finite random samples. These gaps do not invalidate the method but need to be addressed before the central claims are fully supported.","major_comments":[{"comment":"For n>10 the manuscript states that robustness against parasitic ZZ interactions can be inferred from the magnitude of the constraint C, but Fig. 5 reports only the ratio C1/C0 and the unperturbed infidelity. The first-order parasitic correction in Eq. (25) is a sum of terms λ_j multiplied by coefficients, and C in Eq. (27) is a sum of squares of those coefficients. A small ratio C1/C0 does not imply that the absolute first-order error is small unless C0 is calibrated against the actual parasitic infidelity, and no such calibration is given for n>10. Please report the absolute constraint, or a normalized error bound, and show that its magnitude is consistent with the exact n≤10 data; otherwise the claim of explicit robust solutions for up to 20 spins in Sec. V.B is an unverified extrapolation.","section":"Section V.B, Fig. 5"},{"comment":"The statement that resampling the ensemble every 50 iterations yields pulses that are robust against any error within the error hypercube is stronger than the evidence. The objective is an average over 60 training samples, and validation uses 1000 random points in a hypercube of dimension n-1 (for example n=30 in Fig. 1). No worst-case bound, Lipschitz estimate, or covering argument is supplied. The 99.9% fidelity claim should therefore be phrased as average robustness over the sampled error distribution, or it should be supplemented by a deterministic certification method.","section":"Section IV, Appendix B"},{"comment":"The objective J in Eq. (3) measures overlap in operator space, not state fidelity. Unless a spectral-gap argument is provided, a small operator infidelity does not directly imply high fidelity of the prepared eigenstate |ψ(T)> to the target state |ψ_T>. The abstract and Sec. I state '99.9% fidelity', which normally denotes state fidelity. Please either prove the connection in the IQC framework or present state-fidelity data for the small systems where exact propagation is available.","section":"Eq. (3)"},{"comment":"The perturbative constraint removes the parasitic term only to first order in λ_j. At the 5% error level used in Fig. 5, second-order corrections may be non-negligible, especially as n grows and the number of parasitic terms increases. The paper should quantify the regime of validity of the first-order treatment, for example by comparing the exact n≤10 parasitic infidelity with the first-order prediction based on Eq. (25).","section":"Section V.A"}],"minor_comments":[{"comment":"The axes labels contain typos: 'Avergage' and 'in-delity' should be corrected.","section":"Fig. 5"},{"comment":"The notation [5%,-5%]^{n-1} is inconsistent with the interval [-Δε, Δε] used in the main text; the orientation of the interval should be fixed.","section":"Appendix B"},{"comment":"The product of propagators in Eq. (B5) has an ambiguous ordering; the time ordering should be specified explicitly.","section":"Appendix B, Eq. (B5)"},{"comment":"The text says the objective function is 'maximized exactly' for the target, while Eq. (3) defines an infidelity that is minimized; the wording should be made consistent.","section":"Sec. II B"},{"comment":"Ref. [46] is cited as 'to appear in PRX Quantum'; the citation should be updated to the published version if available, and otherwise the dependence of the closed-algebra construction on that work should be stated clearly.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The core method is plausible and the small-system verification is a genuine strength. The main obstacle is the gap between the verified n≤10 results and the n=20 headline in the parasitic case, compounded by the finite-sample nature of the robustness claim. If the authors can supply absolute constraint values, a normalization argument, or additional evidence for n>10, and if they sharpen the fidelity/robustness wording, I would support acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is a solid extension of implicit quantum control (IQC): it adds ensemble-averaged robustness against coupling fluctuations and a perturbative constraint for parasitic ZZ interactions outside the closed subalgebra. The numerical work is substantial — cluster-state and GHZ-state protocols for chains up to 30 qubits, exact verification for the parasitic case up to n=10, and release of code and pulse data. The quadratic-scaling operator formalism is used intelligently, and the robust pulses show clear improvements over non-robust ones. That part deserves credit.\n\nThe soft spots are in the strength of the claims. The abstract promises \"explicit control solutions for spin chains with parasitic interactions of up to 20 spins\" and \"99.9% fidelity\" under 5% coupling errors. But the n>10 parasitic claim is supported only by the ratio C1/C0 in Fig. 5, not by absolute values of the constraint or by any calibration connecting C to actual parasitic infidelity. A small ratio means only that the optimized pulse is better than the unoptimized one; it does not bound the first-order error unless C0 itself is known to be small. That is a missing link. Separately, the \"99.9% fidelity\" number comes from the operator infidelity of Eq. (3), which is not the same as state fidelity. The paper does not prove the two are close for these non-projector operators, so the headline metric is ambiguous. Finally, robustness over the whole error hypercube is inferred from 1000 random validation samples; that is a reasonable practical check but not a guarantee, and the wording \"entire 5% interval\" is stronger than the evidence.\n\nThese are not fatal flaws. The method is plausible, the exact n≤10 verification supports it, and the gaps are addressable with additional analysis — for instance, showing absolute C1 values, testing a few worst-case points, or proving a bound connecting Eq. (3) to state fidelity. The paper is clearly written and the citations look appropriate, including the companion IQC paper from the same group.\n\nWho should read it: people working on scalable optimal control, especially for measurement-based quantum computing and sensing with spin chains. It deserves a serious referee, but the referee should insist on closing the parasitic extrapolation and clarifying the fidelity metric before the claims are accepted.","headline":"Extends implicit quantum control to robust ensemble optimization and parasitic terms, with real numerical evidence up to 30 qubits, but the headline claims outrun the verification: the n>10 parasitic robustness is inferred from a ratio, not an absolute error bound.","tokens_in":15907,"tokens_out":2721,"would_cite":true,"duration_ms":27315,"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":"This paper claims that robust optimal control of spin chains can be done with polynomial cost, achieving 99.9% fidelity for 30-spin cluster states and 20-spin chains under 5% coupling errors.","keywords":["robust quantum control","implicit quantum control","spin chains","cluster states","GHZ states","parasitic ZZ interaction","perturbative control","quantum sensing"],"falsifier":"Using the published optimized pulses for a 10-spin cluster state, set every coupling error to the hypercube corner $\\epsilon_j = +\\Delta\\epsilon$ so that all couplings are 5% too strong, propagate the Schrödinger equation exactly, and compute the infidelity; if it exceeds $10^{-3}$, the claimed robustness over the entire 5% error interval fails.","tokens_in":14964,"feed_emoji":"⚛️","tokens_out":7347,"duration_ms":64873,"temperature":0.7,"pith_summary":"Robust quantum control usually becomes numerically prohibitive as the system grows, because simulating the controlled dynamics costs exponential effort in the number of qubits. This paper tries to remove that bottleneck: it represents the evolving state implicitly as an eigenstate of an operator that obeys the von Neumann equation, so the relevant dynamics live in a space of quadratic size in the chain length. With this representation, the authors construct single control pulses that prepare a 30-spin cluster state and a GHZ-based sensing protocol with infidelity below $10^{-3}$ even when every spin-spin coupling is off by up to 5%. They also treat parasitic ZZ couplings, which break the closed-commutator structure, through a first-order perturbative condition that suppresses their leading effect and extends to chains of up to 20 spins. If the claims hold, noise-resilient control becomes available for interacting many-body systems well beyond the few-qubit regime.","feed_headline":"Robust control pulses reach 30-spin chains","feed_subtitle":"Implicit operator method keeps cluster-state and GHZ preparation accurate even with 5% coupling errors.","key_machinery":"The machinery is the implicit-operator representation: $I(t)|\\Psi(t)\\rangle = \\gamma|\\Psi(t)\\rangle$ with $I(t)$ evolving by $\\dot I(t) = -i[H(t), I(t)]$. For the chain Hamiltonian the operator basis $\\{a_j\\}$ has $2n^2+3n+1$ elements, so the adjoint equation is a $d$-dimensional linear system with $d$ quadratic in $n$. Robustness comes from averaging the infidelity over an ensemble of Hamiltonians with different coupling values and sampling the average gradient over random points. For terms outside the closed basis, the key identity is $Q(\\tau) = U_0^\\dagger(t,t-\\tau) H_P U_0(t,t-\\tau)$, whose equation of motion can be integrated efficiently; since products of solutions are also solutions, the parasitic ZZ term is propagated as $U_0^\\dagger Z_j U_0\\, U_0^\\dagger Z_{j+1} U_0$, leading to the first-order cancellation condition $C=0$.","core_discovery":"The paper's central claim is that robust control of interacting spin chains does not require an ensemble average over the exponentially many noise configurations or full-state simulation. A single time-dependent Hamiltonian, obtained by minimizing the ensemble-averaged infidelity over 60 random error samples and validated on 1000 fresh samples, keeps average infidelity around $10^{-3}$ for $n=30$ cluster-state preparation at $\\Delta\\epsilon/g = 5\\%$; the same strategy improves GHZ-state preparation and the final measurement mapping by several orders of magnitude. For parasitic ZZ interactions, the paper introduces a perturbative measure $C$ whose vanishing removes the first-order effect of every parasitic strength $\\lambda_j$ simultaneously, and shows that pulses satisfying it stay accurate up to $n=20$ chains.","pith_inferences":["The paper leaves implicit that the perturbative cancellation should also work for any parasitic term that factorizes into operators from the closed basis; testing $X_j X_{j+1}$ or $Y_j Y_{j+1}$ parasitic terms would delimit how general the mechanism is.","A reader should not assume the sampling-based robustness proof covers the full error hypercube; evaluating the published pulses at boundary configurations such as all $\\epsilon_j = +\\Delta\\epsilon$ would convert the statistical evidence into a worst-case statement.","Because the first-order condition removes the effect of every $\\lambda_j$ at once, a natural further step is to combine the constraint with higher-order perturbative corrections to extend the robustness beyond the small-$\\lambda$ regime."],"forward_implications":["Cluster states for measurement-based quantum computing can be prepared on 30 spins with infidelity near $10^{-3}$ using pulses whose duration grows linearly with $n$ and whose amplitudes stay of order $g$.","A complete Heisenberg-limited sensing protocol, including both entangled-state preparation and the final readout mapping, can be made robust against 5% coupling disorder, with four- and three-order-of-magnitude infidelity improvements respectively.","Parasitic ZZ interactions, a dominant error in superconducting platforms, can be suppressed to first order for any unknown strength, and the suppression ratio $C_1/C_0$ grows only moderately with system size up to 20 spins.","Mean infidelity below $10^{-3}$ can be obtained with 60 training samples, so the ensemble size needed for robust control need not grow exponentially with the number of fluctuating couplings."],"supporting_citations":[{"why":"Supplies the implicit-control framework and the quadratic operator basis that makes the polynomial scaling possible.","marker":"[46]"},{"why":"Provides the random-sampling argument used to replace the exponential ensemble average over coupling errors.","marker":"[51]"},{"why":"Justifies treating any non-degenerate eigenstate of $I(t)$ as a solution of the Schrödinger equation, the foundational identity of the method.","marker":"[47]"},{"why":"Define the cluster state as a resource for measurement-based quantum computation, the target of the first control task.","marker":"[52, 53]"},{"why":"Establishes the Heisenberg limit that motivates the GHZ-based sensing application.","marker":"[55]"},{"why":"Documents the parasitic ZZ interaction in superconducting qubits that motivates the perturbative correction.","marker":"[56]"},{"why":"Support the noise model in which two-qubit coupling errors are larger than single-qubit errors.","marker":"[49, 50]"}],"fun_headline_variants":["Implicit control reaches 30-spin chains","No-state-simulation control hits 30 spins","Robust 30-spin cluster prep without Hilbert space","Operator-only control tames 30-spin chains"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that one pulse works for every error configuration in the $\\pm5\\%$ interval rests on the assumption that the 60 training points and the 1000 validation points represent the worst case in that continuous hypercube.","fun_headline_variants_meta":{"raw":{"variants":["Implicit control reaches 30-spin chains","No-state-simulation control hits 30 spins","Robust 30-spin cluster prep without Hilbert space","Operator-only control tames 30-spin chains"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1228,"prompt_tokens":762,"completion_tokens":466,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":378,"completion_tokens_details":{"reasoning_tokens":404}},"tokens_in":378,"tokens_out":466,"duration_ms":4599,"temperature":1.0,"reasoning_tokens":404,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T20:29:53.905688+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Using the published optimized pulses for a 10-spin cluster state, set every coupling error to the hypercube corner $\\epsilon_j = +\\Delta\\epsilon$ so that all couplings are 5% too strong, propagate the Schrödinger equation exactly, and compute the infidelity; if it exceeds $10^{-3}$, the claimed robustness over the entire 5% error interval fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the implicit-control framework and the quadratic operator basis that makes the polynomial scaling possible."},{"cited_title":"Krinner, P","cited_arxiv_id":null,"evidence_quote":"Provides the random-sampling argument used to replace the exponential ensemble average over coupling errors."}],"review_version":1}