{"id":"4376acef-9d11-4f33-b535-4cfeaef2f124","arxiv_id":"2508.20887","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A multi-ansatz variational quantum linear solver, tested on a quantum simulator, reproduces classical 1D shock-tube solutions and benefits from additional ansatz branches and domain decomposition.","lead":"This paper describes a variational quantum linear solver that adds multiple, independently-parameterized quantum circuits to solve the linear systems coming from implicit time stepping of the compressible Navier-Stokes equations. It shows simulator results on 1D shock tubes where the multi-ansatz approach tracks classical solutions, and reports that more ansatz branches and domain decomposition improve convergence.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Cost function in Eqs. 14–15 is not the squared residual; minimizing the written cost does not yield the normalized solution, so the solver as described may not solve Ax=b.","rationale":"The reader's weakest assumption identifies the same algebraic inconsistency in Eqs. 14–15, and I agree that this is the most load-bearing concern. The central claim depends on the solver converging to the solution of Ax=b; if the minimized cost is not the squared residual or a functionally equivalent proxy, the good-looking shock-tube plots cannot be reproduced. The paper actually displays a correct-looking cost in Eq. 16 (the standard VQLS ratio), but does not reconcile it with Eq. 14, leaving serious ambiguity. My additional counterexample shows the two costs disagree even at the solution and that residual minimization over normalized states is not safe. This does not necessarily refute the empirical results—the authors may have silently used the correct cost—but it invalidates the derivation and demands correction. Other issues (missing norm recovery, lack of quantitative errors, no code) exist but are secondary; this cost-function inconsistency is concrete, internal, and directly threatens the claimed solution. The reader's CONDITIONAL verdict remains appropriate: the paper can be accepted only after the cost function is correctly derived, the implemented cost is explicitly stated, and code/data are provided for reproduction.","tokens_in":11808,"tokens_out":12352,"duration_ms":124226,"concrete_test":"Run a minimal numerical check: choose A=diag(2,1), b=(1,0), whose exact normalized solution is x*=(1,0). Optimize (e.g., by fine grid search) the cost from Eq. 14 and the ratio cost from Eq. 16 over normalized two-qubit states. If the minimizer of Eq. 14 is not x*, the written cost is invalid. Then, if the authors' simulation code is released, instrument the optimizer to log the exact cost expression used; if it is the ratio form of Eq. 16, the derivation needs amending but the results may be reproducible. This single check settles whether the solver's objective matches the intended linear system.","verdict_should_be":"UNCHANGED","load_bearing_attack":"For the central claim to hold, the VQLS must minimize a cost whose global minimum is the (normalized) solution of Ax=b. The standard VQLS cost is C = 1 − ||²/(||A|x>||²||b||²), zero iff A|x> ∝ |b>. The paper instead writes Eq. 14 as E = ||A|x> − |b>||² and claims it is equivalent to Eq. 15, E = . These are not equivalent: for normalized |x>, Eq. 14 = − 2Re() + 1, while Eq. 15 = − ||². A concrete counterexample, A=diag(2,1), b=(1,0), x=(1,0), gives Eq. 14 = 1 and Eq. 15 = 0. Moreover, minimizing the residual squared over normalized states does not generally return the normalized solution to Ax=b: for the same example, the minimizer of ||A|x>−b||² subject to ||x||=1 is not x*=(1,0). Thus if the implemented cost is Eq. 14, the solver is minimizing the wrong objective. If instead the implemented cost is the ratio in Eq. 16 (which is the correct VQLS cost), then the derivation in Eqs. 14–15 is erroneous and must be corrected. Either way, the manuscript fails to provide a valid, self-consistent cost function, and the numerical results cannot be trusted without code or a precise statement of which cost was optimized.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a hybrid quantum-classical solver for the linear systems that arise from implicit discretization of the one-dimensional compressible Navier-Stokes equations. The core idea is a variational quantum linear solver (VQLS) whose trial state is a classically-weighted superposition of outputs from multiple parameterized quantum circuits ('multi-ansatz tree'). The method is tested on three one-dimensional shock-tube configurations using an 8-qubit quantum virtual machine, with classical LU decomposition as the reference. The authors report that the solver reproduces shock, rarefaction, and contact discontinuities, and that increasing the number of ansatz trees and domain-decomposition blocks improves convergence and stability. The paper also discusses scalability and NISQ compatibility.","tokens_in":12306,"tokens_out":4285,"duration_ms":44727,"significance":"If the central claims hold, the paper would be a useful step toward integrating variational quantum linear solvers into CFD workflows, particularly through the multi-ansatz tree architecture and domain-decomposition strategy for mitigating barren-plateau and expressivity limitations. The use of an external classical LU baseline is a methodological strength, and the authors are honest about observed artifacts in the most demanding case. However, the current manuscript contains a load-bearing algebraic inconsistency in the definition of the cost function, and the numerical evidence is presented almost entirely qualitatively. These issues must be resolved before the claims can be accepted.","major_comments":[{"comment":"The claimed equivalence between Eq. (14) and Eq. (15) is incorrect. For normalized |x> and |b>, Eq. (14) expands to − 2Re() + 1, while Eq. (15) expands to − ||². These differ by a cross-term. A concrete counterexample: A = diag(2,1), |b>=(1,0), |x>=(1,0) gives Eq. (14)=1 and Eq. (15)=0. More seriously, minimizing the squared residual over normalized states does not in general return the normalized solution to Ax=b; the correct VQLS cost function is the ratio form in Eq. (16). The manuscript must state unambiguously which cost function is actually minimized. If Eq. (16) is the implemented cost, then Eqs. (14)-(15) must be corrected or removed; if Eq. (14) is implemented, the solver is minimizing the wrong objective. This is central to the validity of all numerical results.","section":"Section 2.3, Eqs. (14)-(15)"},{"comment":"The accuracy claims are not quantified. The text states that the quantum solver reproduces the classical solution 'with high fidelity' and 'quantitative agreement' for Cases 1 and 2, but no error norms, L2 errors, maximum deviations, or convergence rates are reported. Figures 5-8 show only overlaid qualitative curves. Since the central claim is accurate capture of discontinuities, quantitative error metrics (e.g., L1/L2 error versus the LU reference, perhaps as a function of iteration count) should be added. Without them, 'accurate' is unsupported.","section":"Section 3.2 and Figs. 5-8"},{"comment":"The pronounced high-frequency artifacts in Case 3 are attributed to 'stochastic fluctuations introduced by quantum noise during measurement and state preparation' and to ansatz expressivity limitations. However, the simulations are performed on a quantum virtual machine, and no noise model is described in Section 3.1. If the QVM is noiseless, the quantum-noise explanation is unsupported; if a noise model is used, it must be specified. Additionally, repeated runs with error bars would be needed to distinguish stochastic effects from deterministic ansatz error.","section":"Section 3.2, Case 3"},{"comment":"The discretization equations are too garbled to be reproducible. Equation (7) has unbalanced parentheses and missing indices; the flux linearization in Eq. (10) appears corrupted; and the definitions of D, U, and L in Eq. (13) are incomplete and ambiguous. Since the matrix A in the linear system is the very object that must be decomposed into unitaries for the VQLS, a clean, unambiguous derivation of the discrete linear system is essential.","section":"Section 2.2, Eqs. (7)-(13)"}],"minor_comments":[{"comment":"The notation W and V is introduced but not explicitly defined in the displayed equation; please clarify that W = A†|b>); the dagger and inner-product ordering should be checked. Also, the text says 'the quantum solution is projected to a classical vector'—please clarify whether the output is a classical vector obtained by full tomography or a quantum state norm.","section":"Section 2.3, after Eq. (16)"},{"comment":"There is a typo in Eq. (1): 'Al' should be 'A†'. In Eq. (14), the expansion should use A†A, not At A, and the missing Hermitian conjugate should be fixed.","section":"Section 2.2, Eq. (1) and Eq. (14)"},{"comment":"The table entries for Reynolds numbers are unclear: 'Re, 0' appears to be a typo, and the definitions of Re_L and Re_R should be stated explicitly. The units for pressure and density are also not defined.","section":"Table 1"},{"comment":"The abbreviation 'QVM' should be defined at first use, and the nature of the quantum virtual machine (noiseless simulation, noise model, shot count, etc.) should be described.","section":"Section 3.1"}],"recommendation":"major_revision","confidential_remarks":"The paper fits the special topic and addresses a relevant problem. The main blocker is the cost-function inconsistency in Eqs. (14)-(15); if the implemented cost is actually Eq. (16), the numerical results may be salvageable, but the written derivation must be corrected and the implemented cost confirmed. I would also require quantitative accuracy metrics and a clarification of the noise model before publication. The literature review includes several works from the same group; this is not disqualifying but should not be the sole basis for the claimed novelty."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the multi-ansatz tree is a reasonable variant of VQLS and applying it to implicit compressible Navier-Stokes on a simulator is a legitimate niche step. The paper is not a breakthrough, but it is not a recycle either. What it does well: the architecture (superposition of several ansatze plus domain decomposition) is clearly described, and the validation against LU on three 1D shock-tube cases is an honest external benchmark. The authors also admit the artifacts in Case 3 instead of hiding them.\n\nThe soft spots are real. The most serious is the cost function. Eq. (14) defines E as the squared residual ||A|x>-|b>||^2. Eq. (15) then claims this equals . Those are not algebraically the same: (14) has a linear cross term 2Re() while (15) has ||^2. A simple counterexample shows they differ. The actual cost that appears later, Eq. (16), is the standard VQLS ratio 1 - ||^2 / , which is the correct objective. So the implementation may well be fine, but the derivation presented in Eqs. (14)-(15) is wrong and the manuscript does not state which of the two was optimized in the simulations. That is a load-bearing inconsistency.\n\nThen there are the smaller problems. The discretization formulas (7)-(13) have obvious typos and missing index terms; no quantitative error metrics appear anywhere—only plots and qualitative descriptions; the hyperparameters (number of trees, blocks, tolerances) are tuned on the same test cases; and no code or data are shared. The complexity comparison with LU is hand-wavy and does not account for state preparation or measurement overhead.\n\nIn balance, the central idea is plausible and the paper could be a useful data point for the quantum-CFD community, but as written it is not reproducible and the core derivation does not hold. I would not cite it in its current form. For peer review: send it out, but the referee's job should be to force a rewrite of Section 2.3, a precise statement of the optimized objective, and a minimal set of error bars and reproducibility artifacts. Without those, the numerical results are not trustable.","headline":"Plausible incremental VQLS-for-CFD idea, but the cost function derivation is internally inconsistent and no code or error metrics are provided; needs major revision.","tokens_in":12696,"tokens_out":3985,"would_cite":false,"duration_ms":37646,"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":"A multi-ansatz variational quantum solver reproduces shock-tube solutions of the compressible Navier-Stokes equations.","keywords":["variational quantum linear solver","multi-ansatz tree","compressible Navier-Stokes equations","shock tube","implicit time integration","barren plateaus","NISQ","quantum computational fluid dynamics"],"falsifier":"Take a small linear system with a known solution, run the multi-ansatz VQLS to convergence, and compare the cost-function value with the directly computed residual ||A|x>-|b>|| from the reconstructed classical state; if they disagree beyond normalization, the objective being minimized is not the residual, and the converged state is not the solution.","tokens_in":11745,"feed_emoji":"⚛️","tokens_out":7667,"duration_ms":74164,"temperature":0.7,"pith_summary":"The paper claims that a variational quantum linear solver (VQLS) built from multiple ansatz branches can solve the linear systems produced by implicit time discretization of the one-dimensional compressible Navier-Stokes equations. On a quantum virtual machine, it reproduces classical shock-tube solutions for shock, rarefaction, and contact discontinuities. The core move is to represent the trial solution as a weighted sum of outputs from distinct parameterized circuits, trained jointly, which enlarges the representable solution space without deepening the circuit. Parametric studies report that increasing the number of ansatz trees and partitioning the domain into blocks both accelerate convergence and improve stability, even with limited qubits. If the reported behavior holds, this gives a NISQ-compatible route to quantum-accelerated CFD that does not require fault-tolerant hardware.","feed_headline":"Multi-ansatz quantum solver reproduces shock-tube flows","feed_subtitle":"Variational circuits plus domain decomposition bring implicit Navier-Stokes linear solves to near-term quantum hardware.","key_machinery":"The multi-ansatz tree: a trial state formed as a linear combination of outputs of several structurally different parameterized circuits, with classical coefficients trained in an outer loop. Its role is to expand the accessible solution space without increasing circuit depth, to diversify gradient directions across the ansatz ensemble, and to permit pruning of low-sensitivity Krylov branches — the mechanism the paper invokes to escape barren-plateau-limited optimization.","core_discovery":"The solver looks for |x> solving A|x> = |b> by preparing a trial state |z> = sum_i a_i |psi(theta_i)> and minimizing the cost C = , whose zero-energy ground state encodes the solution. Each |psi(theta_i)> comes from a different unitary ansatz circuit, and the coefficients a_i are optimized classically alongside the circuit parameters. The linear system itself is (D+U+L) Delta Q = RHS arising from a first-order flux-vector-split implicit scheme for the one-dimensional Navier-Stokes equations. On an 85-cell uniform grid (8 qubits), the method reproduces the classical density, velocity, pressure, and energy profiles in the baseline shock-tube case, with small osci","pith_inferences":["The ansatz-tree construction is not tied to the shock-tube equations; it should transfer to other implicit discretizations (incompressible flows, heat conduction, elliptic systems) that produce an Ax=b step at each timestep.","The pruning idea in the Krylov tree suggests an online adaptive strategy—start with few branches and add only those with high gradient sensitivity—which the paper describes as a principle but does not implement as a dynamic rule.","On real NISQ hardware, gate and measurement noise will set a floor on attainable residual; the high-frequency oscillations in the high-Mach case give a natural baseline for evaluating error-mitigation techniques.","A direct check that the optimized state also solves the original linear system, not merely that the reported cost has stagnated, would validate the method independently of the residual formula used in training."],"forward_implications":["Hybrid quantum-classical CFD can replace the dominant linear-solve step in implicit schemes with a VQLS that runs on near-term quantum hardware.","Increasing the number of ansatz branches and partitioning the spatial domain into blocks are two resource-efficient levers to make convergence tractable under limited qubit counts.","The same ansatz-tree construction can be dropped into other implicit PDE discretizations whenever a linear system Ax=b appears at each timestep.","The method is a quantum-accelerated classical CFD loop rather than a full quantum simulation, since the solution state must be projected back to a classical vector after convergence.","The claimed asymptotic scaling suggests a quantum advantage only for sufficiently large systems and with a suitable classical optimizer and preconditioning strategy."],"supporting_citations":[{"why":"Defines the quantum linear systems problem and the speedup target that the variational solver is designed to meet without fault tolerance.","marker":"[22]"},{"why":"Supplies the variational-eigenvalue-solver framework—parameterized circuits plus classical optimization—on which the VQLS cost minimization is built.","marker":"[27]"},{"why":"Argues for variational quantum algorithms in CFD and establishes the ansatz-performance context the multi-ansatz tree addresses.","marker":"[31]"},{"why":"Demonstrates a previous quantum-hardware simulation of compressible flow, the closest benchmark the paper compares its approach against.","marker":"[35]"},{"why":"Frames the three central challenges (state encoding, nonlinear non-unitary evolution, statistics extraction) that the implicit-VQLS approach is positioned to address.","marker":"[42]"},{"why":"Proposes a quantum lattice Boltzmann method as an alternative quantum CFD path, giving the contrast case for the implicit linear-solve approach.","marker":"[45]"}],"fun_headline_variants":["Multi-ansatz quantum solver tackles shock-tube flows","Quantum variational solver with multi-ansatz tree reproduces shock dynamics","Hybrid quantum-classical solver handles compressible flow linear systems","More ansatz branches improve quantum flow solver convergence","Quantum virtual machine runs multi-ansatz solver for 1D shocks"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The method assumes the minimized cost is exactly the squared residual of the linear system, so the state found at convergence is the solution; if the cost and residual disagree, the solver could converge to a state that does not solve Ax=b.","fun_headline_variants_meta":{"raw":{"variants":["Multi-ansatz quantum solver tackles shock-tube flows","Quantum variational solver with multi-ansatz tree reproduces shock dynamics","Hybrid quantum-classical solver handles compressible flow linear systems","More ansatz branches improve quantum flow solver convergence","Quantum virtual machine runs multi-ansatz solver for 1D shocks"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00031,"raw_usage":{"total_tokens":1611,"prompt_tokens":757,"completion_tokens":854,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":501,"completion_tokens_details":{"reasoning_tokens":769}},"tokens_in":501,"tokens_out":854,"duration_ms":7697,"temperature":1.0,"reasoning_tokens":769,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T14:43:43.565142+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a small linear system with a known solution, run the multi-ansatz VQLS to convergence, and compare the cost-function value with the directly computed residual ||A|x>-|b>|| from the reconstructed classical state; if they disagree beyond normalization, the objective being minimized is not the residual, and the converged state is not the solution.","supporting_citations":[],"review_version":1}