{"id":"b4098a53-a65a-4ca0-b06a-0c765ea05815","arxiv_id":"2608.07149","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Correlated disorder in transmon qubit arrays preserves the many-body localization transition, with a local memory parameter locating the critical point consistently with entanglement entropy analysis.","lead":"This paper studies whether many-body localization, a quantum phenomenon that prevents thermalization, survives in superconducting qubit arrays when the random disorder is correlated rather than fully independent. It finds the transition still occurs at roughly the same disorder strength, and proposes a simple qubit-population measurement that can detect it.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The robustness conclusion is under-supported: the effective Hamiltonian used for Figs. 2-3 is not specified to include the correlated couplings, and no same-model uncorrelated control is shown; the comparison to Ref. [17] is indirect.","rationale":"The reader's conditional verdict already captures part of this concern through the request to simulate the uncorrelated counterpart of the paper's own model. My stress-test sharpens it: the simulation Hamiltonian is under-specified, so it is not yet established that the correlated disorder is present in the numerical data at all. The SM's assertion that weak correlated off-diagonal disorder is 'naturally included' is not a substitute for reporting the actual coupling distribution used in the eigenstate and dynamics calculations. If the couplings were fixed, Figs. 2-3 would test only a standard XY chain with on-site disorder, and the central claim about robustness to correlations would be untested. If they were sampled from the physical flux-dependent expression, the lack of a same-model uncorrelated control still prevents the 'close correspondence' conclusion from being drawn from the data. I do not claim the result is false; I claim the presented evidence is not sufficient to establish it as stated. The GitHub repository makes the proposed two-version test feasible. Because the reader already set the verdict to CONDITIONAL and identified related missing comparisons, I recommend keeping the verdict unchanged rather than escalating to UNVERDICTED or REJECT, pending the repository check and the control simulation.","tokens_in":11058,"tokens_out":12357,"duration_ms":128313,"concrete_test":"Run the exact-diagonalization and memory simulations of Figs. 2 and 3 in two versions: (A) the physical correlated model, with h_m = omega_m(phi_m) relative to the mean and J_nm = g_nm(phi_n, phi_m) for phi_n sampled uniformly in [0, phi_0] with the reported alpha; (B) the uncorrelated control, with the same h_m but J_nm replaced by the clean mean J(phi_0). If Var(S2) peaks and M crossings differ beyond error bars, robustness fails; if they coincide, the claim should be explicitly restricted to weak coupling disorder. Additionally, compute Var(S2) for version B using the paper's own code rather than citing Ref. [17], and compare the extracted critical points.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that correlated disorder leaves the MBL transition close to the uncorrelated case. The numerical evidence must therefore come from a Hamiltonian that actually contains the correlation derived in Eq. (2). As written, the effective model in the Supplemental Material (SM Eq. S2) only specifies h_m in [-Delta J, Delta J], with 'J' a characteristic energy scale; it never states whether the couplings J_nm are sampled from the flux-dependent physical expression g_nm(phi_n, phi_m) or fixed at their mean values. If they are fixed, the simulated model is the standard XY chain with independent on-site disorder, and the correlated-disorder part of the claim is never tested. The SM says 'weak correlated off-diagonal disorder' is 'naturally included,' but no coupling-distribution parameters (alpha, phi_0, realization ensemble) for the main runs are reported. Furthermore, the uncorrelated benchmark is not obtained from the same model: Ref. [17] is an external frustrated XY chain, so the agreement at Delta approximately 2 could reflect different couplings or different disorder distributions rather than robustness to correlations. The conclusion's 'fully consistent' therefore overstates what the presented data demonstrate.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the MBL transition in linear transmon arrays whose qubit frequencies and couplings both depend on local external fluxes. The authors derive a covariance between on-site frequency fluctuations and coupling fluctuations (Eq. 2), show that this covariance can be amplified by choosing the junction asymmetry and reference flux (Fig. 1), and then use exact diagonalization for L=8-14 and time evolution for L=8-16 to extract critical disorder strengths from the variance of the half-system 2nd-order Rényi entropy and from a newly introduced local memory parameter. Their central conclusion is that correlated disorder, which is unavoidable in flux-tunable transmon devices, leaves the MBL transition in close correspondence with the uncorrelated case, with a critical dimensionless disorder near Δ≈2.","tokens_in":11214,"tokens_out":5615,"duration_ms":57416,"significance":"If substantiated, the robustness claim is practically important: it would mean that flux-tunable transmon arrays can be treated with standard diagonal-disorder MBL predictions even though the same flux fluctuations also modulate the couplings. The paper is also useful in proposing experimentally accessible probes (Rényi entropy via randomized measurements and single-site populations for the memory parameter), and in providing a reproducible GitHub repository for the processed data. The physical derivation of the covariance is straightforward, and the numerics are standard for finite-size MBL studies. However, the central claim requires the simulations to actually include the correlated couplings and to be compared with a same-model uncorrelated control; the manuscript as written does not establish either condition.","major_comments":[{"comment":"The effective disordered Hamiltonian in SM Eq. (S2) is written with generic hoppings J_nm and on-site terms h_m ∈ [-ΔJ, ΔJ], but the paper never states whether the J_nm used in Figs. 2 and 3 are drawn from the flux-dependent physical couplings g_nm(φ_n,φ_m) of Eq. (S4) or are fixed at their mean values. If the couplings are fixed, the simulated model is just the standard XY chain with independent diagonal disorder and the correlated-disorder claim is never tested; if they are sampled, the coupling-fluctuation amplitudes, the reference flux φ0, and the realization ensemble must be specified. The SM statement that 'weak correlated off-diagonal disorder' is 'naturally included' is not a substitute for reporting the actual distribution used in the numerics.","section":"Supplemental Material, Eq. (S2)"},{"comment":"The robustness claim is compared only to Ref. [17], an external frustrated XY chain, not to the uncorrelated version of the same transmon model. To conclude that 'correlated disorder leaves the MBL transition in close correspondence with the uncorrelated case,' the authors should simulate the same Hamiltonian with independent on-site disorder and fixed (or uncorrelated) couplings, and report the critical disorder estimates for both ensembles. Without this control, the agreement at Δ≈2 may reflect differences in the model or disorder distributions rather than robustness to correlations.","section":"Conclusion and Entanglement signatures"},{"comment":"The internal consistency of the two diagnostics is asserted rather than quantified: the derivative peak of the memory parameter at L=16 occurs at Δ*≈1.82, while the Rényi variance peak for L=14 occurs at (Δφ)_max≈2.17 and the von Neumann result is ≈2.41 (SM Fig. S2b). The manuscript calls these 'fully consistent' without finite-size extrapolations, error bars, or a scaling collapse. Since the central quantitative claim is that the correlated transition occurs at essentially the same disorder strength as the uncorrelated one, the spread of roughly 20-30% among the extracted points needs to be addressed.","section":"Quantum Dynamics, Fig. 3"}],"minor_comments":[{"comment":"The main text uses Δφ both for the random flux fluctuation and for the dimensionless disorder parameter Δ(φ0)=δ(φ0)/J(φ0) defined in SM Eq. (S12); the notation should be distinguished in Figs. 2-3 and throughout.","section":"Main text and SM, Eq. (S12)"},{"comment":"The title says 'MBL Induced by Correlated Disorder', while the conclusion claims robustness of MBL against correlations; the title may overstate or mislead about what is demonstrated.","section":"Title"},{"comment":"In Figs. 1(b,c), define ω1 and g12 explicitly, and state whether the plotted quantities are dimensionless and what flux distribution is used for the 3,000 realizations.","section":"Fig. 1 caption"},{"comment":"The Hamiltonian in Eq. (1) includes sums over all pairs m<n, but Fig. 1(a) depicts a linear chain; clarify which couplings are retained in the simulations and how the network-mediated non-local terms enter the numerical model.","section":"Eq. (1) and Fig. 1(a)"},{"comment":"The Github repository is a helpful resource; consider archiving the exact version of the scripts in a permanent repository, given that full production runs require HPC resources and are not included.","section":"Data Availability"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the mapping from transmon circuit parameters (junction asymmetry, flux, capacitance participation) to a disordered XY model whose on-site fields and couplings are correlated because they share the same flux fluctuations. That part is clearly derived and useful for anyone designing MBL experiments on superconducting hardware. The numerical work is standard exact diagonalization and time evolution, but it's competently done: three diagnostics (Rényi entropy, von Neumann entropy, memory parameter) line up around Δ≈2, and they ship code and processed data in a GitHub repo, which is more than many theory papers do.\n\nThe soft spots are mostly about what the simulations actually contained. The Supplemental Material defines the effective Hamiltonian H_dis with h_m in [-ΔJ, ΔJ] but never explicitly says whether the J_nm in the runs are sampled from the flux-dependent physical expression or fixed at their mean. The text claims weak correlated off-diagonal disorder is \"naturally included,\" and Fig. S1e–h shows coupling fluctuations, so my guess is they did include them—but a referee shouldn't have to guess. If the couplings were fixed, the paper's central robustness claim would only be about on-site disorder, which would be a more minor result. The paper also never states which α and φ_0 were used for Figs. 2–3, so the main results are not fully reproducible from the text alone. And the \"consistency with the uncorrelated case\" rests on an indirect comparison to Ref. [17], a frustrated XY chain, rather than a direct simulation of the same model with independent on-site disorder. That's a fixable weakness: one additional curve per figure would settle it.\n\nThe other caveats are standard for the MBL numerical literature: no error bars on disorder-averaged quantities, and the memory parameter is evaluated at a single time τ_L≈864 without a convergence check. These are minor and don't sink the paper.\n\nWho gets value: people planning MBL experiments on transmon arrays and anyone who wants a concrete recipe for generating correlated disorder in circuit-QED models. It's not a breakthrough, but it's a useful, honest numerical study. I'd send it to a serious referee, with the request that the authors clarify the model and add the direct control. The code makes this easy to check.","headline":"Useful hardware-to-disorder mapping, clean numerics, but the model actually simulated is under-specified and the robustness claim needs a direct uncorrelated control.","tokens_in":11826,"tokens_out":4328,"would_cite":true,"duration_ms":40729,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that in strongly coupled transmon chains, the correlated disorder produced by external flux noise leaves the many-body localization transition essentially unchanged, with the critical dimensionless disorder strength…","keywords":["many-body localization","transmon qubits","correlated disorder","Rényi entropy","local memory parameter","XY chain","superconducting circuits"],"falsifier":"A tractable calculation would be to simulate the transmon-derived XY Hamiltonian with the correlated off-diagonal fluctuations artificially amplified by a factor of ten so that they become comparable to the frequency disorder; if the variance peak of the half-system Rényi-2 entropy shifts noticeably from Δ≈2.17 or the memory-parameter crossing moves, the central robustness claim is falsified. Alternatively, an experiment on a transmon chain with CC/CS ≈ 0.2 measuring the entropy-variance peak location as a function of applied flux would settle the question directly.","tokens_in":10791,"feed_emoji":"🧊","tokens_out":5270,"duration_ms":47602,"temperature":0.7,"pith_summary":"The paper asks whether many-body localization (MBL) survives in superconducting transmon arrays when disorder is correlated, as it inevitably is when random local fluxes shift both qubit frequencies and qubit-qubit couplings at once. It answers yes: by tuning junction asymmetry and reference flux, the correlations can be amplified, yet both the block-entanglement variance and a new local-memory parameter locate the MBL transition at roughly the same dimensionless disorder strength as the uncorrelated case. The paper also introduces a dynamics-based memory observable that requires only single-qubit population measurements, making the transition experimentally accessible without full state tomography. A sympathetic reading is that the transmon platform's intrinsic disorder correlations do not spoil its ability to host a localized phase, which matters for processor design.","feed_headline":"Correlated disorder preserves many-body localization in qubit chains","feed_subtitle":"Flux-induced coupling noise barely shifts the MBL critical point, easing design of localized transmon processors.","key_machinery":"The argument runs through an effective spin-1/2 XY Hamiltonian whose on-site frequencies ω_m(φ) and couplings g_nm(φ) both inherit randomness from local fluxes φ_n. The dimensionless disorder parameter Δ(φ0)=δ(φ0)/J(φ0) maps hardware control (maximum flux φ0) onto the theory's disorder axis, and the covariance amplification coefficient A_m(φ0) quantifies how strongly coupling fluctuations track frequency fluctuations. Two diagnostics carry the numerical case: the disorder-ensemble variance of half-system Rényi-2 entropy, whose finite-size maximum marks the critical point, and a normalized local-memory parameter M(t) built from the profile distance D(t), whose late-time value crosses near Δ≈2 and whose derivative peaks sharpen with L.","core_discovery":"The central claim is that correlated disorder in a transmon array leaves the MBL transition in close correspondence with the uncorrelated case, indicating that the transmon platform is robust against correlations in its random couplings and fields. This is established by exact diagonalization of an effective spin-1/2 XY chain derived from the transmon circuit: the disorder-ensemble variance of half-system Rényi-2 entropy peaks at a dimensionless disorder strength of about 2.17 for the largest size studied (L=14), and a newly introduced local-memory parameter, computed from the long-time dynamics of site occupations, shows a scale-invariant crossing near Δ≈2 with a derivative peak that sharpens with system size. The paper further shows that engineering the junction asymmetry and reference flux controls the covariance between frequency and coupling fluctuations, providing a hardware-level dial for correlated disorder.","pith_inferences":["If the same robustness holds for next-nearest-neighbor and network-mediated couplings beyond the linear chain studied here, then MBL in two-dimensional transmon lattices would inherit the same insensitivity to correlated noise; the paper's mapping of flux-controlled disorder to an effective XY model gives a route to test this numerically.","The claim suggests a practical design principle: rather than engineering away correlated disorder, processors can tune junction asymmetry to operate at a flux sweet spot where correlations are weak enough that localization physics is unchanged.","A sharper test would be to measure the frequency-coupling covariance directly via two-qubit spectroscopy across many flux realizations; if Cov(ω_m, g_nm) scales as the paper's linearized expression, the effective model's premise is confirmed.","The finite-size critical estimates (Δ≈2.17 from entropy variance, Δ≈1.82 from memory derivative) leave a gap that presumably closes with L; extrapolating both to the thermodynamic limit would test whether correlated and uncorrelated transitions truly coincide."],"forward_implications":["A transmon chain with strong capacitive coupling (CC/CS ≈ 0.2) can be operated in a localized phase whose transition point is set by the same dimensionless disorder Δ≈2 seen in uncorrelated models, so existing MBL design rules transfer to realistic devices.","The local-memory parameter M(t) gives an experimentally cheap probe of MBL: only site occupations are needed, not entanglement tomography, and its crossing point agrees with entropy-based criticality.","Flux control can engineer correlated disorder patterns in situ; by choosing junction asymmetry α and reference flux φ0 the covariance amplification coefficient A_m(φ0) can be made large or small.","Mid-spectrum eigenstates show an area-law-to-volume-law crossover via the Rényi-2 entropy variance, with the variance maximum growing with system size, so finite devices up to L=14 already show the precursor of the MBL transition."],"supporting_citations":[{"why":"It supplies the uncorrelated disordered frustrated XY chain that defines the benchmark, including the dimensionless disorder parameter and finite-size critical estimates that the correlated-disorder results must match.","marker":"[17]"},{"why":"It establishes that the disorder-ensemble variance of block entanglement peaks at the MBL transition and is the diagnostic the paper adapts to Rényi-2 entropy.","marker":"[12]"},{"why":"It provides the transmon Hamiltonian and the flux-dependent Josephson energy E_J(φ)=E_J,0[cos^2(πφ)+α^2 sin^2(πφ)]^{1/2} used to engineer correlated disorder.","marker":"[25]"},{"why":"It derives the mediated, beyond-nearest-neighbor capacitive couplings that make non-local interactions non-negligible at CC/CS=0.2.","marker":"[19]"},{"why":"It shows that Rényi entropies can be estimated from randomized measurements, justifying the choice of S_2 as an experimentally accessible probe.","marker":"[27]"},{"why":"It demonstrates Rényi-entropy measurement on a superconducting processor, supporting the experimental feasibility of the entropy diagnostic.","marker":"[18]"}],"fun_headline_variants":["Correlated noise doesn't break many-body localization in qubits","Qubit arrays stay localized despite correlated disorder","MBL survives correlated disorder in transmon networks","Disorder correlations don't shift qubit localization transition","Transmon MBL robust to correlated flux noise"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire comparison to the uncorrelated benchmark assumes that flux-induced fluctuations in the qubit-qubit couplings are substantially smaller than the frequency fluctuations, so that the dimensionless disorder Δ=δ(φ0)/J(φ0) is a faithful single-parameter description; the supplemental states this but provides no quantitative bound.","fun_headline_variants_meta":{"raw":{"variants":["Correlated noise doesn't break many-body localization in qubits","Qubit arrays stay localized despite correlated disorder","MBL survives correlated disorder in transmon networks","Disorder correlations don't shift qubit localization transition","Transmon MBL robust to correlated flux noise"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000461,"raw_usage":{"total_tokens":2302,"prompt_tokens":935,"completion_tokens":1367,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":551,"completion_tokens_details":{"reasoning_tokens":1292}},"tokens_in":551,"tokens_out":1367,"duration_ms":9339,"temperature":1.0,"reasoning_tokens":1292,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T13:57:21.348126+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A tractable calculation would be to simulate the transmon-derived XY Hamiltonian with the correlated off-diagonal fluctuations artificially amplified by a factor of ten so that they become comparable to the frequency disorder; if the variance peak of the half-system Rényi-2 entropy shifts noticeably from Δ≈2.17 or the memory-parameter crossing moves, the central robustness claim is falsified. Alternatively, an experiment on a transmon chain with CC/CS ≈ 0.2 measuring the entropy-variance peak location as a function of applied flux would settle the question directly.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the uncorrelated disordered frustrated XY chain that defines the benchmark, including the dimensionless disorder parameter and finite-size critical estimates that the correlated-disorder results must match."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It establishes that the disorder-ensemble variance of block entanglement peaks at the MBL transition and is the diagnostic the paper adapts to Rényi-2 entropy."},{"cited_title":"Yanay, J","cited_arxiv_id":null,"evidence_quote":"It derives the mediated, beyond-nearest-neighbor capacitive couplings that make non-local interactions non-negligible at CC/CS=0.2."},{"cited_title":"Elben, B","cited_arxiv_id":null,"evidence_quote":"It shows that Rényi entropies can be estimated from randomized measurements, justifying the choice of S_2 as an experimentally accessible probe."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It demonstrates Rényi-entropy measurement on a superconducting processor, supporting the experimental feasibility of the entropy diagnostic."}],"review_version":1}