{"id":"06f16afb-77a6-4288-8591-f911c39fb0f6","arxiv_id":"2412.12871","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A circuit using two vacuum ancillas and six SUM gates maps a continuous-variable state so that measuring the ancillas in momentum samples its Husimi Q-function, enabling single-shot coherent-state estimation, Gaussian parameter recovery, and Q-tomography.","lead":"The paper defines a quantum coherent state transform and shows a six-gate circuit that makes a continuous-variable state's position and momentum information readable from two ancilla modes. Measuring the ancillas samples the Husimi Q-function, which the authors use for single-shot gate calibration, Gaussian state estimation, Q-function tomography, and CV-DV state transfer.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The exact QCST theorem is not supported by the displayed circuit: the last gate in FIG. 1 must be e^{i p1 p3/(2 sqrt(2))}, not e^{i p1 q3/(2 sqrt(2))}, for Eq. (8)'s reset factorization to hold.","rationale":"The paper's central constructive claim is Theorem 1: six SUM gates implement QCST(|psi>)|0>_c. The first half of the circuit is sound: the symmetric product e^{i q1 q3/sqrt(2)} e^{i sqrt(2) q2 p3} e^{i q1 q3/sqrt(2)} indeed equals e^{i sqrt(2)(q1 q3 + q2 p3)} and produces the coherent-state decomposition of Eq. (6), so HQS sampling is correct. The load-bearing step is the reset: Eq. (8) claims the last three gates equal e^{i(p1 p3 - p2 q3)/sqrt(2)}. The displayed gates are A = e^{i p1 p3/(2 sqrt(2))}, B = e^{-i p2 q3/sqrt(2)}, and C = e^{i p1 q3/(2 sqrt(2))}. Since B and C are not in the required symmetric A-B-A form, a direct BCH computation shows the product is not the claimed unitary; it contains an uncancelled i p1 q3/(2 sqrt(2)) term. Replacing C by A gives the exact reset. The error is localized and easily testable, and HQS plus the single-shot estimation applications are unaffected because the last gates commute with p1 and p2. The other noted weaknesses, such as the Heisenberg-limit language and missing numerical code/error bars, are secondary and do not change the conditional assessment.","tokens_in":13400,"tokens_out":39984,"duration_ms":349943,"concrete_test":"Symbolically evaluate U = exp(i p1 q3/(2 sqrt(2))) exp(-i p2 q3/sqrt(2)) exp(i p1 p3/(2 sqrt(2))) with p1=u, p2=v as c-numbers. Show the BCH result contains an uncancelled i u q3/(2 sqrt(2)) term, so U is not exp(i(u p3 - v q3)/sqrt(2)) and does not map |(u+iv)/2> to |0>. Then replace the p1 q3 factor by exp(i p1 p3/(2 sqrt(2))) and verify the product equals exp(i(p1 p3 - p2 q3)/sqrt(2)) and maps |alpha> to |0>. This separates a label typo from a genuine circuit error.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1 asserts in Eq. (8) that the last three SUM gates multiply to e^{i(-p2 q3 + p1 p3)/sqrt(2)}, but the gate list under FIG. 1 is e^{i p1 p3/(2 sqrt(2))}, e^{-i p2 q3/sqrt(2)}, e^{i p1 q3/(2 sqrt(2))}. Treating p1 and p2 as c-number eigenvalues u and v, the time-ordered BCH product of these three gates is exp[i u p3/(2 sqrt(2)) - i v q3/sqrt(2) + i u q3/(2 sqrt(2)) + i u v/8 - i u^2/16], which is not the claimed exp[i(u p3 - v q3)/sqrt(2)]. The missing factor is exactly the second p1 p3 gate: if the final gate is changed to e^{i p1 p3/(2 sqrt(2))}, the symmetric product e^{A} e^{B} e^{A} collapses to exp[i(p1 p3 - p2 q3)/sqrt(2)]. Thus, as printed, the six-gate circuit does not implement QCST(|psi>) |0>_c exactly. The HQS sampling claim survives because all three reset gates commute with p1 and p2, leaving the marginal momentum probabilities unchanged; but the exact-state claim and any protocol relying on vacuum reset, such as CV-DV transfer, inherit this unverified step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a quantum coherent state transform (QCST), implemented by a six-gate circuit with two vacuum ancilla oscillators, which is claimed to map |0>_c|0>_c|ψ> to QCST(|ψ>)|0>_c. Measuring the first two oscillators in the momentum basis is shown to yield samples distributed according to the Husimi Q-function. The authors apply this primitive to single-shot coherent-state parameter estimation, beam splitter and rotation gate calibration, Gaussian state estimation, Husimi Q-function tomography, and approximate CV-to-DV and DV-to-CV state transfer using discrete-variable ancillas.","tokens_in":13702,"tokens_out":15877,"duration_ms":141809,"significance":"If the construction is correct, the Husimi Q-function sampling (HQS) part is an elegant and potentially useful circuit-level primitive: it converts the non-commuting position and momentum information of a CV state into commuting momentum information of two ancillas, and the measurement distribution is exactly the Husimi Q-function. The first-half circuit calculation is self-contained, analytic, and parameter-free, and the numerical demonstrations for Gaussian estimation and Q-function tomography support the practical appeal. The exact QCST statement, however, is not supported by the displayed circuit because of a gate-algebra inconsistency, and there are normalization errors in the central equations. These are local and fixable, but they affect the exact-state claim and the CV-DV reset application, so the paper needs revision before the full set of claims can be accepted.","major_comments":[{"comment":"The gate list under FIG. 1 ends with e^{i p1 q3/(2√2)}, but Eq. (8) claims that the last three gates multiply to e^{i(-p2 q3 + p1 p3)/√2}. These are not equal. With the displayed final gate, the product of the last three gates contains an extra p1 q3 term and does not reset the third oscillator to vacuum. The correct reset gate is e^{i p1 p3/(2√2)}, used symmetrically around e^{-i p2 q3/√2}. As printed, the six-gate circuit does not implement QCST(|ψ>)|0>_c exactly. The HQS sampling probabilities are unaffected because the last three gates commute with p1 and p2, but the exact QCST theorem and any protocol relying on exact vacuum reset, such as CV-DV transfer, are not currently supported. Please correct the figure and the proof consistently.","section":"Proof of Theorem 1, Eq. (8) and FIG. 1"},{"comment":"There is a normalization-factor error in the definition of QCST and in the final line of Eq. (6). Converting dp1 dp2 = 4 d^2α in the preceding line gives 2/√π ∫∫ ⟨α|ψ⟩ |2 Re α>_p |2 Im α>_p |α>_c d^2α, not 2√π times the same integral. The factor 2√π should be 2/√π in both Eq. (1) and the last line of Eq. (6). As written, the state in Eq. (6) is not normalized and is internally inconsistent with the first line of that equation. The probability densities are unaffected by this overall factor, but the exact QCST definition and theorem statement need the corrected normalization.","section":"Eq. (1) and Eq. (6)"},{"comment":"The second component of the inversion formula appears to be missing parentheses or has a sign error. With Eq. (10), a direct calculation gives α β'* - α' β* = -i sin(θ/2) e^{iφ} (|α|^2 + |β|^2), so the correct estimator for the second component should be i(α β'* - α' β*)/(|α|^2+|β|^2). The printed expression i α β'* - α' β* divided by the same denominator is not equal to sin(θ/2) e^{iφ}; for example, for α=β>0 and φ=0, it vanishes instead of giving sin(θ/2). Since this formula is the basis for the claimed single-shot beam splitter gate calibration, it must be corrected before that application is reliable.","section":"Single-Shot Beam Splitter Calibration, Eq. (11)"}],"minor_comments":[{"comment":"The paper calls Δθ = O(α^{-1}) a Heisenberg limit, but with the mean photon number of the initial coherent states scaling as N ≈ 2|α|^2, this is the standard quantum limit scaling Δθ = O(N^{-1/2}), not the usual Heisenberg scaling O(N^{-1}). The definition in Appendix C is nonstandard and should be clarified to avoid misleading readers.","section":"Conclusion and Appendix C"},{"comment":"The abstract says the protocol estimates the parameters of any Gaussian state, while Eq. (12) is written only for pure Gaussian states. The text later says the mixed-state case follows from the Gaussian form, but the equations should be stated for general Gaussian states or the pure-state restriction should be made explicit.","section":"Gaussian State Estimation, Eq. (12) and surrounding text"},{"comment":"The notation I_DV ⊗ |0><0|_CV in the error definition is not explained; clarify what D describes and what the projector represents after the approximate transfer.","section":"Appendix E, Eq. (E3)"},{"comment":"There are several typographical issues, including 'eignebasis' for 'eigenbasis' in the main text and garbled axis labels in FIGS. 2-6. These should be cleaned up in the final version.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The core HQS idea is sound and the main theorem can likely be repaired by correcting the final gate in FIG. 1 (and the corresponding gate in the supplemental figures) and the normalization factors. The beam splitter inversion formula also needs a straightforward fix. The nonstandard use of 'Heisenberg limit' should be addressed, but it does not affect the fundamental HQS construction. The paper is within the scope of the journal and merits consideration after these corrections."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper gives you a clean circuit for sampling the Husimi Q function: two ancilla oscillators, three two-mode SUM gates, momentum measurements. That part works, and it is a genuinely simple way to see that heterodyne statistics can be produced without a mixer. The first three gates are enough; the last three are meant to reset the third oscillator.\n\nThe soft spot is real: the gate list under Fig. 1 does not match the reset factorization in Eq. (8) unless the final gate is e^{i p1 p3/(2√2)} rather than e^{i p1 q3/(2√2)}. With the printed gate, the product of the three reset gates has extra terms, so the circuit is not exactly QCST. The HQS sampling still survives because the reset gates commute with p1 and p2, so the marginal probabilities are unchanged. But the exact-state claim and the CV-DV transfer protocol depend on the unverified step. This looks like a typo rather than a fundamental flaw, but a referee needs to confirm.\n\nThere is also a spelling typo in the normalization: Eq. (6) and the definition of QCST in Eq. (1) have 2√π where the algebra gives 2/√π. The measurement statistics come out right if you use the correct constant, so this is cosmetic but should be fixed.\n\nThe Heisenberg-limit discussion is overblown. In Appendix C they compute ΔH = |α|/√2 and then claim Δθ ~ 1/|α| is the Heisenberg limit. But the photon number is N ~ 2|α|², so the scaling is N^{-1/2}, the standard quantum limit. Same for the rotation gate. The single-shot variance of 1/2 per quadrature is exactly what heterodyne detection gives; that is not a new capability.\n\nWhat is new is the circuit-level implementation and the applications: single-shot gate calibration for beam splitters and rotations, Q tomography via MLE on samples, and an approximate CV-DV transfer scheme. Those are worth exploring. The numerical sections are a bit thin—no code or error bars, and the DV-ancilla convergence is supported only by numerics, not a proof.\n\nOverall: the core idea is simple and likely correct after typo fixes. It deserves a serious referee, not a desk reject. I'd send it back to the authors to correct the gate and the Heisenberg language, then accept if the revisions hold.","headline":"Useful HQS circuit primitive, but the exact QCST claim is undercut by a load-bearing gate typo and the Heisenberg-limit claim is SQL in photon number.","tokens_in":14272,"tokens_out":7994,"would_cite":true,"duration_ms":64448,"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":"The paper introduces the quantum coherent state transform (QCST), a six-gate circuit that encodes a continuous-variable state's position and momentum into two ancilla oscillators, whose momentum measurement samples the Husimi Q-function…","keywords":["quantum coherent state transform","continuous-variable quantum computing","Husimi Q-function","coherent-state POVM","single-shot parameter estimation","gate calibration","CV-DV state transfer","two-mode SUM gates"],"falsifier":"Take the fourth through sixth gates in the printed order, apply them to the state on the right-hand side of Eq. (6), and compute the overlap of the third mode with vacuum; if this overlap is not $1$ for, say, $|\\psi\\rangle = |1\\rangle_F$, or if the third mode retains entanglement with the ancillas, the exact reset step and hence the exact QCST statement fail, even though the momentum-measurement probabilities would still be Q-samples.","tokens_in":13136,"feed_emoji":"⚛️","tokens_out":8229,"duration_ms":69411,"temperature":0.7,"pith_summary":"The paper introduces the quantum coherent state transform (QCST), a unitary map that encodes the position and momentum information of a continuous-variable (CV) quantum state into the momentum amplitudes of two ancilla oscillators. It proves that a circuit of six two-mode SUM gates plus two vacuum ancillas implements this map exactly, and that measuring the ancillas in the momentum basis produces outcomes $\\alpha = (p_1 + i p_2)/2$ whose probability density is the Husimi Q-function of the input state. A sympathetic reader would care because this turns coherent-state sampling into a single primitive: one copy of a coherent state suffices to estimate its parameter at the minimum-uncertainty limit, the same samples calibrate beam splitter and rotation gates at Heisenberg-limited precision, and repeated samples give Q-function tomography or Gaussian parameter estimation. The protocol also admits a DV-ancilla version that approximates QCST and supports CV-to-DV and DV-to-CV state transfer.","feed_headline":"Six-gate circuit samples any CV state's Husimi Q-function","feed_subtitle":"Two vacuum ancillas and six SUM gates turn a single copy of a CV state into coherent-state measurement samples.","key_machinery":"The load-bearing object is the QCST circuit: two vacuum CV ancillas, six two-mode SUM gates, and momentum-basis measurement of the first two modes. The first three gates create the entangled state whose momentum readout is the Q-function; the last three gates are claimed to reset the third oscillator to vacuum via $U_{\\mathrm{reset}} = e^{i(-\\hat{p}_2\\hat{q}_3+\\hat{p}_1\\hat{p}_3)/\\sqrt{2}}$, using the displacement identity $|\\alpha\\rangle\\langle\\alpha| = \\frac{1}{\\pi} \\iint e^{-(p^2+q^2)/2 - 2i(p\\,\\mathrm{Re}\\,\\alpha + q\\,\\mathrm{Im}\\,\\alpha)} e^{i\\sqrt{2}(p\\hat{q}+q\\hat{p})}\\,dq\\,dp$ obtained from the Wigner–Weyl transform of the coherent state's Gaussian Wigner function. The key mechanism is that momentum readout converts the non-commuting $(\\hat{q},\\hat{p})$ information of the input into commuting momentum outcomes of two ancillas, with the reset gates intended to return the input mode to vacuum.","core_discovery":"The central discovery is Theorem 1: the quantum circuit in FIG. 1, acting on two vacuum ancilla oscillators and an arbitrary CV state $|\\psi\\rangle$, implements $|0\\rangle_c |0\\rangle_c |\\psi\\rangle \\mapsto \\mathrm{QCST}(|\\psi\\rangle) |0\\rangle_c$, where $\\mathrm{QCST}(|\\psi\\rangle) = \\frac{2}{\\sqrt{\\pi}} \\iint \\langle \\alpha | \\psi \\rangle |2\\,\\mathrm{Re}\\,\\alpha\\rangle_p |2\\,\\mathrm{Im}\\,\\alpha\\rangle_p \\, d^2\\alpha$. Measuring the two ancilla modes in the momentum eigenbasis and setting $\\alpha=(p_1+ip_2)/2$ yields a probability density equal to the Husimi Q-function $Q(\\alpha)=\\frac{1}{\\pi} \\langle\\alpha|\\rho|\\alpha\\rangle_c$. Equivalently, the measurement is a coherent-state POVM with elements $\\{\\frac{1}{\\pi} |\\alpha\\rangle\\langle\\alpha|\\}$, so one circuit run draws a sample from the Q-function of the unknown state. The paper further claims this sample stream enables single-shot coherent-state parameter estimation at the minimum-uncertainty limit, single-shot Heisenberg-limited calibration of beam-splitter and rotation gates, Gaussian state parameter estimation from sample moments, efficient Q-function tomography by maximum likelihood, and approximate CV-DV state transfer with DV ancillas.","pith_inferences":["The paper does not spell this out, but the QCST circuit can be read as a circuit-level coherent-state heterodyne measurement, so the same device that samples $Q(\\alpha)$ could serve as a subroutine inside larger hybrid circuits for quantum state discrimination or Bayesian phase estimation.","The DV-ancilla approximation error in Eq. (E3) is governed by the lattice spacing $N\\lambda$ and the finite square in phase space captured by the grid; the paper leaves open the design rule for choosing $(N,\\lambda)$ optimally for a target state's known support.","The paper needs a separate phase-estimation protocol for displacement gates rather than direct coherent-state parameter estimation, suggesting that QCST samples directly certify gates whose action preserves coherent states up to parameter-dependent displacement or rescaling, and extending it to more general gates is an open direction.","The arbitrary-angle momentum readout mentioned in Appendix A could be combined with QCST to sample rotated Husimi functions, connecting it to standard homodyne tomography without pointwise scans; the paper does not develop that generalization."],"forward_implications":["One copy of an unknown coherent state $|\\beta\\rangle$ is enough to estimate $\\beta$ with variance $1/2$ in each quadrature, the minimum-uncertainty limit, and choosing bright inputs $|\\alpha\\rangle$ with large $\\alpha$ makes beam-splitter and rotation gate calibration errors $O(|\\alpha|^{-1})$.","With $M$ samples from a Gaussian state, the sample mean and covariance of $\\alpha$ estimate the Q-function's mean and covariance with error $O(M^{-1/2})$, allowing squeezing-gate parameters to be calibrated.","For non-Gaussian states, repeated runs of the same circuit plus maximum-likelihood reconstruction recover the Husimi Q-function with error scaling $O(M^{-1/2})$, without pointwise phase-space scans.","The same construction extends to multimode states: applying the circuit per mode lets all $2n$ ancilla momentum outcomes sample the joint Husimi Q-function.","Replacing CV ancillas with DV systems and conditional displacement gates approximates QCST; in the limit of fine and wide grids the CV state is reset to vacuum and the information is transferred to DV registers, giving CV-DV and DV-CV state transfer."],"supporting_citations":[{"why":"Defines the hybrid oscillator-qubit processor instruction set and the two-mode SUM gates the circuit is built from.","marker":"[1]"},{"why":"Supplies the SUM-gate convention in the GKP-code setting used by FIG. 1.","marker":"[25]"},{"why":"The existing single-shot displacement-gate interferometry protocol that QCST extends to beam-splitter and rotation gate calibration.","marker":"[12]"},{"why":"Establishes the Heisenberg limit scaling used to certify the gate-calibration error bounds.","marker":"[14]"},{"why":"Gives the pointwise Q-function tomography protocol that the HQS plus maximum-likelihood method is compared against.","marker":"[20]"},{"why":"Provides the Gaussian state learning framework and sampling complexity used for Gaussian parameter estimation.","marker":"[24]"},{"why":"The CV-to-DV state transfer protocol that the DV-ancilla QCST approximates and compares with.","marker":"[26]"},{"why":"The quantum AD/DA conversion framework used for inverse DV-to-CV transfer and Fourier transform tasks.","marker":"[27]"}],"fun_headline_variants":["Six-gate QCST samples any CV state's Q-function","Two ancillas, six SUM gates: single-copy CV state estimation","Single-shot Heisenberg-limited calibration via QCST","QCST: from one CV copy to Husimi Q-function samples","Six SUM gates turn a CV state into coherent-state samples"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the last three SUM gates factor into the single unitary $e^{i(-\\hat{p}_2\\hat{q}_3+\\hat{p}_1\\hat{p}_3)/\\sqrt{2}}$ that resets the third oscillator to vacuum; the paper states this factorization without derivation, and the printed gate list shows a final gate $e^{i\\hat{p}_1\\hat{q}_3/(2\\sqrt{2})}$ that does not match the required $e^{i\\hat{p}_1\\hat{p}_3/(2\\sqrt{2})}$, so the exact QCST output depends on an unproved gate-algebra step.","fun_headline_variants_meta":{"raw":{"variants":["Six-gate QCST samples any CV state's Q-function","Two ancillas, six SUM gates: single-copy CV state estimation","Single-shot Heisenberg-limited calibration via QCST","QCST: from one CV copy to Husimi Q-function samples","Six SUM gates turn a CV state into coherent-state samples"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000346,"raw_usage":{"total_tokens":1979,"prompt_tokens":1111,"completion_tokens":868,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":727,"completion_tokens_details":{"reasoning_tokens":783}},"tokens_in":727,"tokens_out":868,"duration_ms":8015,"temperature":1.0,"reasoning_tokens":783,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:43:21.481485+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the fourth through sixth gates in the printed order, apply them to the state on the right-hand side of Eq. (6), and compute the overlap of the third mode with vacuum; if this overlap is not $1$ for, say, $|\\psi\\rangle = |1\\rangle_F$, or if the third mode retains entanglement with the ancillas, the exact reset step and hence the exact QCST statement fail, even though the momentum-measurement probabilities would still be Q-samples.","supporting_citations":[{"cited_title":"Progress towards practical qubit computation using approximate gottesman-kitaev- preskill codes","cited_arxiv_id":null,"evidence_quote":"Supplies the SUM-gate convention in the GKP-code setting used by FIG. 1."},{"cited_title":"Single-shot quantum signal pro- cessing interferometry","cited_arxiv_id":null,"evidence_quote":"The existing single-shot displacement-gate interferometry protocol that QCST extends to beam-splitter and rotation gate calibration."},{"cited_title":"Quantum-enhanced measurements: beating the standard quantum limit","cited_arxiv_id":null,"evidence_quote":"Establishes the Heisenberg limit scaling used to certify the gate-calibration error bounds."},{"cited_title":"Quantitative tomography for continuous variable quantum systems","cited_arxiv_id":null,"evidence_quote":"Gives the pointwise Q-function tomography protocol that the HQS plus maximum-likelihood method is compared against."},{"cited_title":"Universal unitary trans- fer of continuous-variable quantum states into a few qubits","cited_arxiv_id":null,"evidence_quote":"The CV-to-DV state transfer protocol that the DV-ancilla QCST approximates and compares with."},{"cited_title":"Quantum Coherent State Transform on Continuous-Variable Systems","cited_arxiv_id":null,"evidence_quote":"The quantum AD/DA conversion framework used for inverse DV-to-CV transfer and Fourier transform tasks."}],"review_version":1}