{"id":"1934422a-787d-4cc5-98bc-1a48124f7426","arxiv_id":"2509.02112","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A quantum state tomography method that estimates Chebyshev coefficients of the encoded function, achieving measurement costs independent of qubit count for smooth, large-scale-dominated functions.","lead":"Researchers propose measuring Chebyshev expansion coefficients of a quantum state's encoded function using a Hadamard test, then reconstructing a truncated approximation. This yields an approximate tomography method whose shot count depends on the function's smoothness rather than the qubit count.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Known-state-preparation assumption undermines 'tomography' framing; method only applies to states with controllable U_ψ, not unknown states.","rationale":"The reader's weakest assumption is exactly the known-U_ψ requirement, and I agree. This is the most load-bearing issue because it changes the scope of the central claim: the method does not perform tomography on unknown states, but rather reads out states produced by known circuits. The complexity scalings (O((4m/3)^d/(ε d!)) and O(m^d/d!)) are independent of n even if ε should be ε², so the qualitative independence claim survives; the controlled-U_ψ limitation, however, is fundamental and cannot be patched by adjusting constants. The paper does credit open-source code, which is positive, but the framing in the abstract and title overstates the generality. The concrete swap-test experiment would settle whether an alternative exists; if it fails, the paper must be reframed as a readout technique, requiring major revision and making conditional acceptance appropriate.","tokens_in":12347,"tokens_out":11598,"duration_ms":128347,"concrete_test":"Use Qiskit to prepare a 6-qubit state |ψ⟩ with a random circuit whose unitary is hidden from the reconstruction routine. Provide the routine only with copies of |ψ⟩ and the Chebyshev basis-preparation circuits. Replace the Hadamard test in Fig. 2 with a swap test to estimate |⟨ψ|T_s,n⟩|², and attempt to reconstruct f(x) by assigning signs via the normalization constraint and smoothness priors. If the reconstructed function is sign-degenerate or the fidelity on analytic test functions (e.g., sin πx) is far below the reported 97.74%, this demonstrates that complex inner products—and hence controlled-U_ψ—are indispensable, confirming the limitation.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section III.A explicitly assumes 'Given state preparation unitaries |ϕ⟩=U_ϕ|0⟩ and |ψ⟩=U_ψ|0⟩', and the Hadamard-test circuit in Fig. 2 implements controlled-U_ψ to create the superposition needed for measuring Re⟨ψ|T_s,n⟩. For a truly unknown state—the standard QST setting—U_ψ is unavailable. A swap test between |ψ⟩ and |T_s,n⟩ estimates only |⟨ψ|T_s,n⟩|², losing the sign/phase needed for the complex coefficient. Alternative interferometric schemes also need a coherence between |ψ⟩ and |T_s,n⟩ that effectively requires knowledge of U_ψ. Thus QST-CP is not a general tomography protocol; it is a readout method for states whose preparation circuit is known and controllable. The abstract's unqualified 'approximate tomography method' and the comparison to 'exponential scaling of full measurement-based QST' are therefore misleading. The central claim is contingent on a restriction that is neither stated as a limitation nor reflected in the title.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes QST-CP, an approximate tomography method for pure quantum states that encode continuous (possibly multivariate) complex-valued functions. The target state is expanded in a truncated Chebyshev basis; the expansion coefficients are obtained as inner products between the target state and Chebyshev-basis states, measured via a Hadamard test. The authors give a circuit for preparing the normalized Chebyshev basis states, a stopping criterion based on the cumulative coefficient energy A_m, and complexity claims stating that the number of measurement repetitions and the classical post-processing are independent of the qubit count n. Numerical simulations on analytic functions and on turbulent-flow data are used to validate the method.","tokens_in":12615,"tokens_out":5519,"duration_ms":65933,"significance":"If the central claims hold, the paper would offer a practical readout method for quantum states produced by known circuits, particularly in quantum computational fluid dynamics, where states encode smooth flow fields. The mathematical core is largely self-contained: the expansion coefficients are defined as discrete inner products and are directly measured, so the procedure is not circular in the sense of fitting parameters from the measured data. The manuscript also provides explicit circuits, a clear stopping criterion, reproducible code, and numerical demonstrations on realistic flow data. However, the protocol's dependence on a known and controllable state-preparation unitary is a major scope restriction, and the complexity analysis contains a shot-count error and an unclear post-processing claim. These issues affect the paper's central efficiency and tomography claims.","major_comments":[{"comment":"The Hadamard-test inner-product measurement requires a controlled version of U_psi for the target state. For a genuinely unknown state—the standard QST setting—U_psi is not available, and a swap test yields only |<psi|T_{s,n}>|^2, losing the sign/phase needed for the complex coefficients. Thus QST-CP is a readout protocol for states whose preparation circuit is known and controllable, not a general tomography method. The abstract and Section II.A compare to full measurement-based QST without flagging this restriction. Please state this limitation explicitly and temper the tomography framing, or provide a sign-recovery scheme that does not require U_psi.","section":"Section III.A, Fig. 2"},{"comment":"The measurement-repetition count is underestimated by a power of epsilon. Estimating an inner product to precision epsilon with a Hadamard test requires O(1/epsilon^2) shots, not O(1/epsilon). Consequently the total repetition count should be O((4m/3)^d/(epsilon^2 d!)) for the multivariate case and O(4m/(3 epsilon^2)) for the single-variable case, not the expressions in the text and Table I. The qualitative n-independence survives, but the stated complexity is incorrect and must be corrected.","section":"Section III.C, Table I"},{"comment":"The post-processing claim is ambiguous and, as stated, inconsistent with reconstructing a quantum state. If the output is a reconstructed state vector of length 2^n, evaluating the truncated Chebyshev series at all grid points costs at least Omega(2^n) (or Omega(2^n m^d) depending on the evaluation method), so the post-processing cannot be O(1) or O(m^d/d!) independent of n. If the output is only the set of expansion coefficients or a single amplitude, that is not full state reconstruction. Please clarify what 'post-processing' computes and reconcile it with the tomography claim.","section":"Section III.C, Table I"}],"minor_comments":[{"comment":"The discrete orthogonality relation is stated for 0 <= s,t <= p, but for s=t=p the sum is zero because T_p vanishes at its own zeros. The correct statement is for 0 <= s,t < p. Since the expansions use P <= p-1, this does not affect the main results, but the displayed equation is formally false.","section":"Eq. (5)"},{"comment":"The notation is inconsistent: the subscript list is written as (s_1,...,s_n,n_1,...,n_d) and (s_1,...,s_n) in the summation in Eq. (13). The number of variables is d, not n. Please correct the subscripting.","section":"Eq. (12)"},{"comment":"All numerical results use only 500 measurement repetitions per coefficient. The visible deviations in Figs. 6(b,d) are attributed to sampling error; a brief note on the expected amplitude of statistical fluctuations at 500 shots would help the reader interpret the fidelity numbers.","section":"Section IV.A"}],"recommendation":"major_revision","confidential_remarks":"The main concern is scope: the method requires a known, controllable preparation unitary for the target state, which is not standard QST. If the authors reposition the contribution as a readout/coefficient-extraction protocol for states produced by known circuits, and fix the epsilon scaling and post-processing clarification, the paper could be publishable. The numerical demonstrations are useful but do not resolve the foundational framing issue."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the Chebyshev coefficient measurement idea is new and the circuit construction is clever, but the paper sells itself as general tomography when it only works for states with a known and controllable preparation unitary. That, plus a wrong scaling in the shot count, means the main efficiency claims need rework.\n\nWhat's actually new: using a Hadamard test to directly estimate Chebyshev expansion coefficients of the encoded function, then truncating based on cumulative energy. The discrete orthogonality setup is clean, and the circuit in Fig. 3 is an elegant way to prepare Chebyshev basis states. The numerical tests on analytic functions and DNS flow fields are honest and show the method captures large-scale features with modest measurements. If the state is produced by a known circuit, this is a practical readout scheme.\n\nThe soft spots are real, though. First, the tomography framing is misleading. The Hadamard test needs controlled-U_ψ, so the target state's preparation circuit must be known and controllable. Standard QST is for unknown states; this is a readout for known circuits. The abstract's comparison to 'exponential scaling of full measurement-based QST' is apples to oranges. That should be flagged as a limitation, not buried in the text. Second, the complexity analysis is off. Section III.C says each basis needs O((4/3)^d/ε) repetitions; a Hadamard test estimate to accuracy ε needs O(1/ε^2) shots. Table I also lists O(1) post-processing for single-variable, but the text says summing m bases, which is O(m). These errors need fixing. Third, the linear-depth claim in the abstract is conditional on U_ψ being shallow; the paper admits this in III.C but the abstract overstates it. The stopping criterion is heuristic, not a rigorous error bound, which is fine but should be worded as such.\n\nOverall: the core idea holds up for a narrower application than the title suggests. It's a solid contribution for QCFD readout, but needs a serious revision before publication. I'd send it to peer review — the idea deserves referee time — but with the expectation of major changes.","headline":"Clever spectral readout for states with known preparation circuits, but the 'tomography' framing overreaches and the shot-count scaling is wrong by a factor of ε.","tokens_in":13051,"tokens_out":2981,"would_cite":true,"duration_ms":32844,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P68","41A10"],"pacs":["03.67.-a"],"model":"deepseek-v4-flash","headline":"This paper makes approximate tomography of function-encoding quantum states far cheaper by measuring only a truncated Chebyshev expansion, with measurement and post-processing costs that do not grow with qubit count.","keywords":["quantum state tomography","Chebyshev polynomials","Hadamard test","amplitude encoding","spectral truncation","fluid dynamics","quantum readout","complexity scaling"],"falsifier":"Run QST-CP on a state encoding a function whose Chebyshev coefficients decay slowly, for example $f(x)=\\operatorname{sgn}(x)$ or a sinusoid whose frequency grows with n, at fixed threshold $A_c=0.9$ and increasing qubit count n. If achieving the threshold forces the stopping order m to grow with n, then the claimed n-independence of measurement repetitions fails outside scale-concentrated states; if m stays constant while fidelity remains high, the claim is universal.","tokens_in":12262,"feed_emoji":"⚛️","tokens_out":10578,"duration_ms":105076,"temperature":0.7,"texified_at":"2026-08-05T20:22:23.693498+00:00","pith_summary":"Full quantum state tomography is exponentially expensive because an n-qubit state has $4^n$ independent parameters. This paper argues that when the state encodes a smooth, physically structured function—say a fluid velocity field—most of that information lives in a few large-scale modes. Its method expands the encoded function in Chebyshev polynomials, measures the expansion coefficients as inner products with Hadamard-test circuits, and reconstructs the state from the truncated series. For such states the measurement repetition count and classical post-processing are independent of the number of qubits, scaling only with the truncation order and spatial dimension. Numerical tests on analytic functions and turbulent flow fields recover dominant structures with fidelities from about 78% to 98% at modest truncation orders.","texify_model":"deepseek-v4-flash","texify_usage":{"total_tokens":6713,"prompt_tokens":774,"completion_tokens":5939,"prompt_tokens_details":{"cached_tokens":0},"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":774,"completion_tokens_details":{"reasoning_tokens":5212}},"feed_headline":"Chebyshev truncation makes quantum readout independent of qubit count","feed_subtitle":"Reconstructing only the dominant scale modes keeps measurement and processing costs flat as systems grow.","key_machinery":"The load-bearing identity is $\\langle T_{s,n} | \\psi \\rangle = \\langle F, \\tilde{T}_{s,2^n} \\rangle_{2^n}$: the overlap between the target state and a Chebyshev basis state equals the discrete Chebyshev inner product, so every expansion coefficient becomes a circuit-measurable quantity. The basis state $|T_{s,n}\\rangle$ is prepared by a quantum-Fourier-transform-style circuit whose $|0\\rangle$ branch carries $\\cos\\left(\\frac{(2k+1)s\\pi}{2^{n+1}}\\right)$ at position $k$, with overall 75% success probability; the Hadamard test with an optional S gate then reads the real and imaginary parts of the overlap. The discrete orthogonality of Chebyshev polynomials at their zeros supplies normalization, and the classical convergence bound for truncated Chebyshev series justifies","core_discovery":"QST-CP reduces approximate tomography of a pure state encoding a continuous function to estimating Chebyshev expansion coefficients. Normalized Chebyshev basis states are constructed so that their inner product with the target equals the discrete Chebyshev coefficient; a Hadamard test with an optional S gate reads real and imaginary parts. The partial sum $A_m$ of squared coefficients tracks captured energy, and measurement stops once $A_m \\geq A_c$. When energy concentrates in low-order modes, m stays small as n grows: measurement repetitions are $O\\left(\\frac{(4m/3)^d}{\\epsilon d!}\\right)$ and post-processing $O\\left(\\frac{m^d}{d!}\\right)$, both independent of n, with linear-depth circuits. Analytic and turbulent-flow tests recover dominant l","pith_inferences":["Extension: the practical reach of the protocol is states whose preparation circuit is known and controllable; for a genuinely unknown state the controlled preparation required by the Hadamard test is not available, so the 'tomography' label applies only in that narrower setting.","Extension: the 75%-success basis preparation contributes the (4/3)^d factor; amplitude amplification on the basis-preparation ancilla would likely reduce this to a constant overhead, improving high-dimensional cases the paper does not explore.","Extension: the same coefficient-estimation loop could be embedded as a readout head in variational or block-encoding solvers, with the solver's own ansatz supplying the controlled preparation unitary and A_m acting as a convergence monitor."],"forward_implications":["Measurement repetitions and post-processing do not grow with qubit count for function-encoding states whose energy is concentrated in low-order modes, directly removing the exponential bottleneck of full tomography in that setting.","The stopping rule A_m ≥ A_c gives a built-in accuracy-efficiency tradeoff: lowering the threshold gives a cheaper coarse reconstruction, while raising it resolves finer scales.","Circuit depth stays linear in n whenever the target state has a shallow preparation circuit, so the method is compatible with near-term hardware for moderate n.","For d-dimensional functions the basis count grows like m^d/d!, so low-dimensional fields are practical; the (4/3)^d factor in repetition count penalizes high-dimensional cases.","The method is tailored to large-scale feature recovery, so it can serve as a readout protocol for quantum PDE solvers whose outputs are smooth fields rather than a replacement for full tomography of arbitrary states."],"supporting_citations":[{"why":"Supplies the Hadamard-test circuit that turns inner-product estimation into a qubit measurement, and the quantum-Fourier-transform structure used to prepare Chebyshev basis states.","marker":"[1]"},{"why":"Supplies the Kolmogorov energy-spectrum decay that motivates the large-scale-dominance assumption for turbulent fields.","marker":"[44]"},{"why":"Supplies the Chebyshev polynomial recurrence, zeros, and discrete orthogonality that define the expansion and the basis states.","marker":"[45]"},{"why":"Supplies the algebraic convergence bound for truncated Chebyshev series, justifying the truncation error control.","marker":"[46]"},{"why":"Establishes the exponential parameter count of full QST and the compressed-sensing route to structured tomography, the baseline this method is contrasted with.","marker":"[26]"},{"why":"Provides the quantum circuit simulation tool used to test the method on analytic functions and flow-field data.","marker":"[47]"},{"why":"Provides the pseudo-spectral method used to generate the 2D isotropic turbulence velocity field used as a test case.","marker":"[48]"},{"why":"Provides the turbulent channel flow dataset used to test reconstruction on 16-qubit encoded fields.","marker":"[49]"}],"fun_headline_variants":["Chebyshev tomography: costs independent of qubit count for smooth states","Quantum state readout that keeps costs flat as qubits grow","Truncated Chebyshev basis makes tomography scale-independent","Quantum tomography with flat cost: Chebyshev truncation","QST-CP: linear-depth tomography with qubit-independent cost"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that the target state has a known preparation unitary that can be controlled for the Hadamard test and that its encoded function concentrates its squared amplitude in low-order Chebyshev modes; if either condition fails, the coefficients either cannot be measured at all or the truncated reconstruction stops being faithful as n grows.","fun_headline_variants_meta":{"raw":{"variants":["Chebyshev tomography: costs independent of qubit count for smooth states","Quantum state readout that keeps costs flat as qubits grow","Truncated Chebyshev basis makes tomography scale-independent","Quantum tomography with flat cost: Chebyshev truncation","QST-CP: linear-depth tomography with qubit-independent cost"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001013,"raw_usage":{"total_tokens":4103,"prompt_tokens":721,"completion_tokens":3382,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":465,"completion_tokens_details":{"reasoning_tokens":3295}},"tokens_in":465,"tokens_out":3382,"duration_ms":27187,"temperature":1.0,"reasoning_tokens":3295,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T11:51:46.887202+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run QST-CP on a state encoding a function whose Chebyshev coefficients decay slowly, for example $f(x)=\\operatorname{sgn}(x)$ or a sinusoid whose frequency grows with n, at fixed threshold $A_c=0.9$ and increasing qubit count n. If achieving the threshold forces the stopping order m to grow with n, then the claimed n-independence of measurement repetitions fails outside scale-concentrated states; if m stays constant while fidelity remains high, the claim is universal.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Hadamard-test circuit that turns inner-product estimation into a qubit measurement, and the quantum-Fourier-transform structure used to prepare Chebyshev basis states."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Kolmogorov energy-spectrum decay that motivates the large-scale-dominance assumption for turbulent fields."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the algebraic convergence bound for truncated Chebyshev series, justifying the truncation error control."},{"cited_title":"Gross, Y","cited_arxiv_id":null,"evidence_quote":"Establishes the exponential parameter count of full QST and the compressed-sensing route to structured tomography, the baseline this method is contrasted with."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the pseudo-spectral method used to generate the 2D isotropic turbulence velocity field used as a test case."},{"cited_title":"Graham, K","cited_arxiv_id":null,"evidence_quote":"Provides the turbulent channel flow dataset used to test reconstruction on 16-qubit encoded fields."}],"review_version":1}