{"id":"c5975d8b-d899-4c12-af03-052204c6e6d1","arxiv_id":"2506.00213","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Disorder in coupling increases localization of two-photon quantum walks, while injection amplitude and phase disorder delocalize them, with correlation nulls showing robustness.","lead":"This paper simulates how fabrication disorder affects two-photon states in nonlinear waveguide arrays. It finds that coupling disorder localizes single-waveguide injection, while injection amplitude and phase disorder delocalize the light, and certain quantum correlations remain robust.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The generalized \"robust null correlations\" claim rests on two special phase configurations; the abstract drops the qualifier, so the main practical takeaway is not yet supported.","rationale":"The paper's core localization trend for coupling disorder is standard Anderson localization and the numerical protocol is plausible, so I do not see that as the weakest point. I also do not treat the m=3 Hilbert-space truncation as load-bearing: with gα0L≈0.0125 the single-pair approximation is excellent, and local dimensions 0,1,2 contain the entire two-photon sector. The genuine soft spot is the gap between the special cases analyzed for correlations and the unqualified abstract statement that null correlations are robust to disorder. A second, independent wording problem is that Sec. III.A explicitly says ballistic propagation 'cannot be affirmed,' while the abstract and conclusions assert a tendency to ballistic propagation; if the authors carry out the proposed phase-sweep test, they should also replace the ballistic sentence with the manuscript's own caveat or supply a power-law/velocity analysis. These are overclaims, not errors in the underlying simulation, so the appropriate disposition remains conditional acceptance pending restriction of the conclusions to the simulated configurations.","tokens_in":10498,"tokens_out":14778,"duration_ms":170308,"concrete_test":"Compute Γ_{q,r} at C0z=20 for phase differences φ_i−φ_p in {0, π/6, π/3, π/2, 2π/3, π} and for κ up to 0.8, recording each disorder realization separately rather than only an ensemble average. If the supposed null elements are exactly zero in all realizations only for φ_i−φ_p=0 and π, or if they become nonzero for generic phases, the abstract must be restricted to the special cases.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing part of the central claim is that the null elements of the correlation matrix are robust to disorder. The evidence is Fig. 5, which examines only two phase configurations (φ_i−φ_p=0 and π), only disorder in the coupling profile C, and a single propagation length C0z=20. The conclusions properly say this is true 'para los casos específicos estudiados,' but the abstract removes that restriction and asserts robustness to disorder in general. Crucially, the paper does not report whether the null entries survive in every disorder realization or only after averaging, nor whether they persist for generic phase differences, for injection amplitude/phase disorder, or at other propagation lengths. If the nulls are a symmetry artifact of the two chosen input phases, the headline claim that fabrication disorder does not disturb biphoton correlations fails in the general form stated. Because this is one of the three positive results advertised in the abstract, the overgeneralization is load-bearing.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports QuTiP simulations of two-photon quantum walks in chi^(2) nonlinear waveguide arrays with disorder imposed on three system parameters: the coupling profile, the injection amplitude, and the injection phase. For single-waveguide injection, coupling disorder is shown to reduce the standard deviation and the participation ratio, indicating increased localization and reduced dispersion. For multi-waveguide injection, amplitude and phase disorder increase the participation ratio, with phase disorder having the stronger effect. The correlation matrix is studied for two input phase configurations, and its null entries appear stable under coupling disorder. The paper concludes that fabrication disorder can localize light while some biphoton correlations remain robust.","tokens_in":10689,"tokens_out":8954,"duration_ms":65183,"significance":"If the central claims hold, the paper provides a useful systematic numerical comparison of three disorder types in an integrated-photonics setting and identifies a potentially interesting robustness property of biphoton correlations. The simulations are direct and parameter-free in the important sense that the Hamiltonian parameters are taken from the experimental literature (Barral et al., 2021) and no fitting is involved; the localization trend for coupling disorder is supported by the presented averaged simulations. The main weaknesses are the overgeneralized abstract claims, the ambiguous relationship between the single-waveguide and multi-waveguide results, and the lack of validation of the single-pair and Hilbert-space truncation assumptions.","major_comments":[{"comment":"The abstract and conclusions state that propagation \"tends to remain ballistic\" for single-waveguide injection with coupling disorder, but Section III explicitly says \"no se puede afirmar que se trate de una propagación balística\" because a more detailed study of the propagation velocity is required. This is a direct contradiction in a headline claim. Please either remove the ballistic statement from the abstract and conclusions or add a quantitative test, such as a linear fit of sigma versus C0z in the pre-reflection regime with confidence intervals.","section":"Abstract / Section III (Quantum random walk)"},{"comment":"The abstract claims that quantum correlations, particularly null elements, are robust to disorder, but the evidence in Fig. 5 is limited to two input phase configurations (phi_i - phi_p = 0 and pi), disorder in the coupling profile only, and a single propagation length C0z = 20. The text itself restricts the conclusion to \"los dos casos particulares estudiados,\" yet the abstract drops this qualifier. The paper also does not report whether null entries survive in individual disorder realizations or only after averaging, nor whether they persist under amplitude/phase disorder or at other propagation lengths. Please qualify the abstract and either add these tests or state the scope explicitly.","section":"Section III (Correlaciones cuánticas) / Abstract"},{"comment":"The load-bearing modeling assumptions are the single-pair, low-intensity condition and the per-mode Hilbert-space truncation m = 3, but no convergence or validity check is provided. The text says \"se asumió que las condiciones son tales que ocurre un único evento no lineal,\" and m = 3 restricts each mode to at most two photons. If higher-order nonlinear events or the truncation alter the dynamics, all localization and correlation measures could change. Please add a convergence check (for example m = 4) or an estimate of the pair-creation probability at the stated parameters to justify the single-pair assumption.","section":"Section II (Configuración de las simulaciones)"},{"comment":"The abstract presents the delocalization result for amplitude and phase disorder as the counterpart of the single-waveguide coupling-disorder result, saying \"cuando se inyecta solamente una guía de onda... Contrariamente, el desorden en la amplitud y fase de inyección tienden a deslocalizar.\" However, the amplitude- and phase-disorder results in Fig. 4 are obtained when more than one waveguide is injected, as Section III states: \"Cuando se inyecta más de una guía de onda.\" Please clarify in the abstract and conclusions that the two statements refer to different injection configurations, or provide single-waveguide results for amplitude/phase disorder.","section":"Abstract / Section III (Tipos de desorden)"}],"minor_comments":[{"comment":"The Hamiltonian sums from j = 1 to N, but the term C_j A_{j+1} A_j^dag requires C_N A_{N+1}, which is undefined; please define C_N = 0 or restrict the sum to j = 1, ..., N-1 and state the boundary condition.","section":"Section I, Eq. (3)"},{"comment":"The diagonal entry of the correlation matrix appears to be written as A_q^{dag 2} A_q^{dag 2}; the normal-ordered detection probability should involve A_q^{dag 2} A_q^2. Please correct the notation.","section":"Section I, Eq. (6)"},{"comment":"The stated sample length L = 80 cm together with C0 = 250 m^-1 gives C0 L = 200, but all simulations are reported up to C0z = 20 and the correlation matrix is evaluated at C0z = 20; either the length should be L = 8 cm or the propagation range is inconsistent.","section":"Section II (Configuración de las simulaciones)"},{"comment":"No error bars or confidence intervals are shown for the disorder averages; please clarify whether the convergence criterion (last vs. penultimate average below 0.75%) is used in place of a standard error, and consider showing error bars on the averaged quantities.","section":"Section III, Figs. 3 and 4"},{"comment":"The spatial average is taken over 10 < C0z < 20, where quantities are said to be near-stationary, but Fig. 3b shows oscillatory behavior at kappa = 0 in this window; please justify this window or demonstrate that the conclusions are insensitive to its endpoints.","section":"Section II (Configuración de las simulaciones)"},{"comment":"The text refers to parameters \"definidos en la Sección II,\" but Eqs. (4)-(6) appear in Section I; please correct the cross-reference.","section":"Section II"},{"comment":"The manuscript does not include a data or code availability statement; for a purely numerical study, a reproducibility statement or release of the simulation scripts would strengthen the paper.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a credible numerical study whose central qualitative trends appear supported by the presented simulations, but the abstract/conclusion mismatches and the overgeneralized robustness claim need to be fixed before publication. I would also encourage the editor to request a data/code availability statement, since the results are entirely simulation-based and the scripts would be straightforward to share."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plain English summary: This is a modest simulation study of SPDC in a 9-waveguide nonlinear array, comparing disorder in coupling, injection amplitude, and injection phase. The core finding — coupling disorder localizes the quantum walk and reduces dispersion, while amplitude and phase disorder delocalize it — is plausible and supported by the averaged simulations. The systematic comparison of three disorder profiles is the real novelty; Anderson localization itself is established and properly cited via Bai et al.\n\nThe paper does a few things well. The parameters come from experimental papers, the averaging (150+ realizations, convergence criterion 0.75%) is honest, and the results section explicitly warns that a ballistic claim cannot be made from the observed linear sigma vs z. That makes the abstract's 'tiende a ser balística' a direct internal contradiction.\n\nThe bigger problem is the correlation robustness claim. Figure 5 shows only two phase configurations (phi_i - phi_p = 0 and pi), only coupling disorder, and only C0 z = 20. The conclusions correctly restrict to 'los casos específicos estudiados,' but the abstract drops that qualifier. Without per-realization survival, other phase differences, or other propagation lengths, the general statement that null correlations are robust to disorder is not supported. This is load-bearing because it is one of the three advertised positive results.\n\nMinor: the m=3 Hilbert space truncation and single-pair assumption are stated but not validated. A quick convergence check in m would close the gap.\n\nOverall: the core physics is likely correct, but the abstract oversells. This deserves a serious referee because the fixes are straightforward: adjust the abstract and either restrict or extend the robustness analysis. I would not cite it in my own work, but it is a fair reading-group candidate to discuss abstract-vs-results discipline. Send it to review.","headline":"Useful but modest simulation study; core localization trend holds, but abstract overclaims ballistic propagation and general robustness of null correlations.","tokens_in":11159,"tokens_out":2455,"would_cite":false,"duration_ms":23248,"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":"Disorder in waveguide coupling pins down two-photon states, while disorder in the input phase spreads them across the array.","keywords":["disorder","nonlinear waveguide arrays","two-photon states","quantum walks","spontaneous parametric down-conversion","participation ratio","correlation matrix","integrated quantum photonics"],"falsifier":"Measure the two-photon correlation matrix and the photon-number distribution at the final propagation distance in a nine-waveguide nonlinear array with controllable coupling disorder; the localization claim fails if $\\sigma$ or $P_R$ increases with $\\kappa$ for single-waveguide injection, and the robustness claim fails if any entry that is zero at $\\kappa=0$ becomes nonzero as $\\kappa$ grows in the two phase configurations studied.","tokens_in":10311,"feed_emoji":"⚛️","tokens_out":10977,"duration_ms":114142,"temperature":0.7,"pith_summary":"This paper asks how fabrication disorder changes the two-photon states that an array of nonlinear waveguides generates through spontaneous parametric down-conversion. Simulating a nine-waveguide chain, it finds that disorder in the inter-waveguide coupling acts as a localizer: for single-waveguide injection, the spread of the photon-number distribution and the participation ratio both fall as disorder grows, and the propagation keeps its ballistic character. Disorder in the input amplitude and phase acts in the opposite direction, delocalizing the wave function, with phase disorder the stronger of the two. The paper further claims that the null entries of the two-photon correlation matrix—the coincidences that never occur—survive coupling disorder, so the qualitative structure of biphoton correlations is stable against fabrication imperfections. If correct, the results identify which disorder type must be controlled in a device and which can be tolerated.","feed_headline":"Coupling disorder localizes two-photon quantum walks","feed_subtitle":"Coupling disorder pins photons down; input-phase disorder spreads the biphoton state across the array.","key_machinery":"The machinery is the interaction-picture momentum operator $\\hat{M} = \\hbar \\sum_j [ C_j \\hat{A}_{j+1} \\hat{A}_j^\\dagger + \\eta_j \\hat{A}_j^{\\dagger 2} ] + \\mathrm{h.c.}$, whose first term moves one photon between neighboring guides and whose second term is the spontaneous parametric down-conversion source that creates photon pairs from the pump, with $\\eta_j = g \\alpha_j$. The simulation evolves $|\\psi(z)\\rangle$ according to $-i\\hbar\\, d|\\psi\\rangle/dz = \\hat{M}|\\psi\\rangle$, then reads off three diagnostics: the normalized photon-number expectation $n_j$, the standard deviation $\\sigma$ of that distribution, and the participation ratio $P_R = (\\sum_j n_j^2)^{-1}$, which runs from $1$ (one fully occupied guide) to $N$ (complete delocalization). The last diagnostic is the correlation matrix $\\Gamma_{q,r}$, whose entries give the two-photon coincidence probability and whose null entries encode forbidden detections. Disorder enters by drawing each coupling, amplitude, or phase from a uniform distribution whose width is controlled by $\\kappa$, following the same scheme used for tight-binding lattices. The stability of $\\Gamma$'s null entries is what carries the claim that biphoton correlation structure tolerates coupling disorder.","core_discovery":"The central claim is that disorder is not one effect but three. When light enters a single waveguide of a nonlinear array, randomizing the coupling constants $C_j$ increases localization and lowers dispersion: both the standard deviation $\\sigma$ and the participation ratio $P_R$ decrease with the disorder strength $\\kappa$, and the wave function stays closer to the injected guide, with weaker reflections at the array boundaries. Propagating the same quantities in $z$ shows an approximately linear growth of $\\sigma$ before boundary reflections, so the motion keeps the ballistic signature of clean waveguide arrays. When all guides are injected and the disorder is placed instead in the input amplitude or phase, $P_R$ rises toward $N$, meaning the wave function delocalizes; phase disorder does so fastest, reaching near-complete delocalization by $\\kappa = 0.5$. For the two phase configurations studied (even–odd phase difference $0$ and $\\pi$), coupling disorder leaves the null entries of the correlation matrix $\\Gamma_{q,r}$ at zero, and for phase difference $\\pi$ the whole matrix is unchanged; for phase difference $0$ only the relative magnitudes of allowed detections shift.","pith_inferences":["The $m=3$ Hilbert space truncation likely makes the reported localization a lower bound: adding higher photon-number components could open extra paths that counteract some of the pinning, an effect the paper does not explore.","The stability of null correlations suggests a practical design heuristic: encode information in forbidden coincidence events, since those survive disorder, rather than in the precise magnitudes of allowed events, which shift with $\\kappa$.","One testable extension is to scan the even–odd phase difference continuously; the paper studies only $0$ and $\\pi$, so the boundary between invariant and merely stable correlation patterns is unknown.","A natural next step is to repeat the same disorder scan for longer arrays or 2D lattices; the nine-guide chain has strong boundary reflections that may amplify or mask the localization trends."],"forward_implications":["For single-waveguide injection, coupling disorder can serve as a controlled knob to confine biphoton propagation without erasing the ballistic signature of the array.","Coincidence measurements that rely on a zero probability of detecting a pair in certain positions will remain reliable under coupling disorder, because the null entries of $\\Gamma_{q,r}$ persist.","Input-phase disorder is the most disruptive of the three; devices that need delocalized, fully spread biphoton states should randomize phases, while devices that need localization should keep phases stable.","With all waveguides pumped, coupling disorder barely changes the participation ratio, so coupling fabrication errors are unlikely to affect the degree of delocalization in that regime.","For the even–odd phase configuration with difference $\\pi$, the entire two-photon correlation matrix is invariant under coupling disorder, not just its zeros."],"supporting_citations":[{"why":"Supplies the interaction-picture momentum operator and the quantum-state engineering framework for nonlinear waveguide arrays.","marker":"(Barral et al., 2020a)"},{"why":"Defines the coupling, amplitude, and phase profiles and the injection controls used for biphoton state synthesis.","marker":"(Barral et al., 2020b)"},{"why":"Earlier study of two-photon Anderson localization in a disordered quadratic waveguide array that this work extends.","marker":"(Bai et al., 2016)"},{"why":"Provides the uniform-disorder implementation and the participation ratio as the localization diagnostic.","marker":"(Karamlou et al., 2022)"},{"why":"Establishes the SPDC and quantum-walk model for arrays of quadratic nonlinear waveguides.","marker":"(Solntsev et al., 2012)"},{"why":"Introduces the two-photon correlation matrix and null-coincidence effects in waveguide lattices.","marker":"(Bromberg et al., 2009)"},{"why":"Shows waveguide arrays support ballistic quantum walks with negligible decoherence, the baseline for the propagation result.","marker":"(Perets et al., 2008)"},{"why":"Provides the numerical solver used to integrate the Schrödinger evolution in the simulations.","marker":"(Johansson et al., 2013)"}],"fun_headline_variants":["Biphoton localization boosted by coupling disorder in arrays","Phase disorder spreads biphotons; coupling disorder confines them","Two-photon correlations survive disorder in nonlinear waveguides","Coupling disorder pins biphotons; input phase disorder scatters"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire simulation assumes exactly one photon pair is created and each guide never holds more than three photons; if stronger pump light or higher photon numbers change the dynamics, the localization and correlation results could differ.","fun_headline_variants_meta":{"raw":{"variants":["Biphoton localization boosted by coupling disorder in arrays","Phase disorder spreads biphotons; coupling disorder confines them","Two-photon correlations survive disorder in nonlinear waveguides","Coupling disorder pins biphotons; input phase disorder scatters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000952,"raw_usage":{"total_tokens":4194,"prompt_tokens":1214,"completion_tokens":2980,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":830,"completion_tokens_details":{"reasoning_tokens":2912}},"tokens_in":830,"tokens_out":2980,"duration_ms":25293,"temperature":1.0,"reasoning_tokens":2912,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:09:25.484219+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the two-photon correlation matrix and the photon-number distribution at the final propagation distance in a nine-waveguide nonlinear array with controllable coupling disorder; the localization claim fails if $\\sigma$ or $P_R$ increases with $\\kappa$ for single-waveguide injection, and the robustness claim fails if any entry that is zero at $\\kappa=0$ becomes nonzero as $\\kappa$ grows in the two phase configurations studied.","supporting_citations":[],"review_version":1}