{"id":"70f84bb6-db86-4536-a3f8-17d9a5aff240","arxiv_id":"2510.10495","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A GQSP-based compiler synthesizes arbitrary bosonic phase gates from conditional-displacement gates on hybrid qubit-oscillator processors and applies them to anharmonic vibronic dynamics of the uracil cation.","lead":"The paper compiles anharmonic molecular potentials into gate sequences for hybrid qubit-oscillator hardware, using a generalized quantum signal-processing scheme that turns pairs of conditional displacements into programmable bosonic phase gates of almost any shape. If it holds up, it gives near-term trapped-ion and circuit-QED devices a general route to simulating molecular dynamics beyond the harmonic approximation.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 5's strip-holomorphy assumption fails for the asymmetric Morse potentials in the uracil case study, so the O(log 1/ε) Fourier-degree claim is unsupported; Fig. 6's d=39, fidelity 0.9229 contradicts the stated ρ=e^π rate.","rationale":"The reader's weakest assumption is exactly this, and I agree with it. The O(log 1/ε) scaling is the paper's headline advantage; if it fails for the Morse potentials used in the case study, the resource table and the 'moderate circuit depth' claim are unsupported, though the method may still work at higher polynomial cost. The two-F/leakage issue in Lemma 2 is also real but secondary: it can be repaired by defining F as the half-time phase and including δ in the error budget. The cleanest decisive test is direct Fourier-coefficient decay. Since this concern reinforces the reader's CONDITIONAL verdict rather than moving it, I recommend no change.","tokens_in":19430,"tokens_out":7040,"duration_ms":63204,"concrete_test":"Take the Table IV ν26 D2 Morse parameters and Δt=0.3 fs, choose L covering the sampled phase space (e.g., L=6; or the L implied by Fig. 6's grid), compute the exact Fourier coefficients a_k=(1/2L)∫_{-L}^{L} exp(iΔt V(Q)) e^{-iπkQ/L} dQ for k up to ~2000, and plot log|a_k| vs k. Exponential decay with slope π/L would validate Lemma 5; algebraic decay or a plateau would confirm the failure. In the same setup, evaluate the sup error of the d=39 partial sum; it should be ~0.08 as Fig. 6 implies, not the ~10^-50 the ρ=e^π bound yields.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is Lemma 5 (Appendix C, Eqs. C7–C10), which gives d = O(log 1/ε) only if g(x)=exp(iΔt V(x)), extended 2L-periodically, has a holomorphic extension to the strip S_σ with |g|≤B. For the case study's Morse potentials (Table IV), this premise is not satisfiable. V^(D2)_26(Q)=9.46894[e^{-0.08653(Q-0.37635)}-1]^2+... has V(-L)-V(L) growing exponentially with L (e.g., >2 eV at L=5 and >15 eV at L=10 for the ν26 D2 well), so the periodic extension is discontinuous at x=±L for any finite L. A discontinuous periodic function cannot have exponentially decaying Fourier coefficients; the coefficients decay only algebraically, so d=O(log 1/ε) fails and the Table I advantage over digital encodings is unsupported. The paper's own unoptimized numeral (d=39, fidelity 0.9229 in Fig. 6) is direct evidence: with ρ=e^π≈23.14, the stated bound would predict error <10^-50, not ~0.077. This is an internal inconsistency, not a stylistic disagreement. The vacuum-state optimized fidelity 0.9999 does not rescue Theorem 2, which needs a uniform gate approximation over all states relevant to the Trotter evolution.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces OQ-GQSP, a compilation method for hybrid oscillator–qubit processors, which uses generalized quantum signal processing (GQSP) with Z-basis conditional-displacement gates to implement nonlinear bosonic phase gates exp(i Δt V(Q)). The main theoretical claims are: (i) Lemma 1 constructs the required complementary signal operators from two CD gates; (ii) Lemma 2 converts such an approximation into a state-dependent nonlinear phase gate by postselection; and (iii) Theorem 2 asserts that a Trotter step of an anharmonic vibronic Hamiltonian costs O(N M' log(1/ε)+N^2 M) CD gates, with success probability (1-δ)^{2M'N}, giving the headline O(log(1/ε)) Fourier degree. The method is applied to a four-state, multimode model of the uracil cation with Morse and quartic potentials, with numerical demonstrations of Wigner functions and reduced two-mode population dynamics.","tokens_in":19642,"tokens_out":8224,"duration_ms":85952,"significance":"If the central complexity claim were correct, OQ-GQSP would be a meaningful step for hybrid CV-DV quantum simulation: it would compile anharmonic molecular potentials without discretizing the vibrational continuum and would achieve polylogarithmic cost in accuracy, linear in the number of modes. Lemma 1 is a clean and checkable construction, and the numerical dynamics simulation, though on a reduced model, is a forward computation with no fitted outcome parameters. However, the key analytic argument for exponential Fourier convergence is invalid for the potentials used in the case study, so the headline resource advantage is not established. The numerical data in Fig. 6 actually contradict the claimed convergence rate. The paper therefore does not currently support its main contribution as stated.","major_comments":[{"comment":"The hypothesis that the 2L-periodic extension of g(x)=exp(iΔt V(x)) admits a holomorphic extension to the strip S_σ with sup |g|≤B is not satisfiable for the case-study potentials. For the Morse potentials in Table IV, e.g., the ν26 D2 well, V(-L)-V(L) grows with L (by several eV already at L≈5), so the periodic extension is discontinuous at x=±L. A discontinuous periodic function cannot have exponentially decaying Fourier coefficients; the coefficients decay only algebraically, and the bound d=O(log(1/ε)) in Theorem 2 is unsupported. The statement after Eq. (C14) that entire potentials such as polynomials and Morse functions allow σ=L is also false: even for f(z)=z^2, exp(iΔt z^2) is unbounded on any horizontal strip, and the periodic extension has a derivative jump. This is load-bearing because Theorem 2 and the resource comparison in Table I both rely on d=O(log(1/ε)).","section":"Appendix C, Lemma 5 / Theorem 2 (Eqs. C7–C10)"},{"comment":"The paper's own unoptimized result, d=39 with fidelity 0.9229 for the ν26 Morse phase gate, is grossly inconsistent with the stated ρ=e^π≈23.14 rate: the Lemma 5 bound would predict an error far below 10^-50 at d=39, not ~0.077. The observed error is consistent with slow algebraic convergence of a discontinuous periodic extension. The vacuum-state-refined angles achieve 0.9999 but are state-dependent; they do not rescue Theorem 2, which requires a uniform gate approximation over all states occurring in the Trotter evolution. This numerical check should be addressed explicitly.","section":"Section IV.A, Fig. 6"},{"comment":"The construction replaces F(U)†F(U) by I on off-target electronic branches (the ≈ in the penultimate line of Eq. C6) without tracking the induced error in the theorem's error budget. Since |F|≤1 and F approximates a phase, the error is controllable by the Fourier truncation error, but the lemma as stated claims an exact state-dependent phase gate. The approximation should be stated as an additional error term and included in the final ε budget; as written, the proof contains an uncontrolled step.","section":"Appendix C, Lemma 2 proof (Eq. C6)"}],"minor_comments":[{"comment":"The appendix presents 'Lemma 1', 'Lemma 2', and then 'Lemma 5' with no Lemmas 3–4; the numbering is confusing and should be made consistent with the main text.","section":"Appendix C numbering"},{"comment":"The caption states that each OQ-GQSP synthesis is 'bounded by the Fourier series approximation error of 0.001', while Fig. 6 reports fidelity 0.9229 for d=39 and the text says the optimized gate reaches 0.9999. It should be clarified which gates (unoptimized or vacuum-refined) are used in the dynamics simulation and how the 0.001 bound is compatible with Fig. 6.","section":"Section IV.B, Fig. 7 caption"},{"comment":"The sentence 'OQ-GQSP can generate arbitrary bosonic phase functions with Fourier series approximation' is too broad. The construction is limited to the Fourier approximation of phase functions whose periodic extension has sufficient smoothness; otherwise convergence is algebraic. This qualification is important given the case-study potentials.","section":"Section III.C after Lemma 1"},{"comment":"The table header for OQ-GQSP lists the Trotter layer count as p=Γt^2ε^{-1}, which seems dimensionally inconsistent with the definition p=⌈Γt^2/ε_Trot⌉ in Eq. (19). Please align notation and define Γ in the table caption.","section":"Section V, Table I"}],"recommendation":"reject","confidential_remarks":"The stress-test concern is valid and is the basis for the recommendation. Lemma 1 is a correct and useful observation, but the central resource claim of the paper—the O(log(1/ε)) Fourier degree for arbitrary analytic potentials—is undermined by the failure of Lemma 5's hypotheses for the Morse potentials in the uracil cation model. This is not merely a missing technical assumption; the provided numerical data contradict the claimed rate. A corrected version would need either a substantially different approximation scheme or a much weaker complexity claim, which would change the paper's contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, what's worth knowing: the OQ-GQSP construction in Lemma 1 is real and independently verifiable. You make a complementary pair of signal operators A=diag(U,I), B=diag(I,U†) from two Z-basis CD gates plus a reusable ancilla, and then plug them into GQSP. That removes the parity restriction of the earlier Z-basis hybrid QSP and lets you synthesize bosonic phase gates directly from Laurent/Fourier coefficients, avoiding the nested Jacobi–Anger expansion. The postselected state-dependent phase gate in Lemma 2 is also clever, and the reduced-model numerics (error below 0.05 over 40 fs) show the method works in a small regime. Credit where due: the paper is clearly written, and the angle-finding and Trotter machinery are standard.\n\nThe soft spots are real. The main one is Lemma 5. The exponential Fourier decay—the source of the O(log 1/ε) degree—requires the 2L-periodic extension of g(x)=exp(iΔtV(x)) to be holomorphic on a strip. For the uracil cation Morse potentials, V(−L) and V(L) differ by several eV, so the periodic extension is discontinuous at the boundary; it cannot be holomorphic on any strip crossing ±L. The proof's conformal mapping step silently assumes 2L-periodicity. So the O(log 1/ε) claim in Theorem 2 and Table I is unsupported for the paper's own case study. The paper's Fig. 6 doesn't rescue it: d=39 gives fidelity 0.9229, nowhere near what the ρ=e^π tail bound would predict if Fourier truncation were the bottleneck. A correct bound for non-periodic potentials would be algebraic in 1/ε, which would erode the headline advantage over digital encodings. This is fixable—choose L or an interval so the phase matches at boundaries, or state honest convergence rates—but it has to be fixed.\n\nSecond, Lemma 2's error budget drops the F†F − I leakage. After postselection you keep F†F on the non-target electronic states, which is I − G†G, not I. The deviation is order δ and needs to be in the bound. Minor, but it's there.\n\nAlso minor: the paper doesn't give the code, the chosen L, or the vacuum-state optimization protocol, so the numerical claims are not fully reproducible as reported.\n\nWho is this for? People working on hybrid CV-DV compilers and quantum simulation of vibronic models. The core construction is worth knowing even if the asymptotic claims need revision. I'd send it to peer review with a request that the authors either prove a convergence bound that applies to their potentials or scale down the claims. I would not cite the O(log 1/ε) result as it stands.","headline":"The CD-gate OQ-GQSP construction is genuinely new and checkable, but the O(log 1/ε) claim rests on a strip-analyticity assumption that the paper's own Morse potentials violate.","tokens_in":20335,"tokens_out":4957,"would_cite":false,"duration_ms":43880,"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":"A compiler for hybrid oscillator-qubit processors synthesizes arbitrary bosonic phase gates in depth O(log(1/ε)), enabling anharmonic vibronic dynamics without discretizing the continuous variables.","keywords":["oscillator-qubit processors","generalized quantum signal processing","vibronic coupling","uracil cation","Morse potential","bosonic phase gates","conditional displacement gates","Trotter simulation"],"falsifier":"Compute the Fourier coefficients a_k of g(x)=exp(iΔt V_Morse(x)) on [-L,L] for the paper's ν26 mode parameters and check whether |a_k| decays as C ρ^{-|k|} with ρ=e^{πσ/L} or only polynomially. If the decay is algebraic, the d=O(log 1/ε) theorem does not apply to this case and the observed fidelity 0.9229 at d=39 would be the expected behavior, not an anomaly.","tokens_in":101,"feed_emoji":"⚛️","tokens_out":4876,"duration_ms":102831,"temperature":0.7,"pith_summary":"The paper introduces a compilation method, oscillator–qubit generalized quantum signal processing (OQ-GQSP), that allows a hybrid processor with one qubit controlling a bosonic mode to implement a nonlinear phase gate exp(i V(Q) Δt) for an arbitrary analytic potential V. The target gate is approximated as a finite Laurent/Fourier series in the displacement operator, and the paper shows, via two lemmas, that generalized quantum signal processing can synthesize this series using conditional-displacement gates and single-qubit rotations. If the construction works as claimed, an anharmonic vibronic Hamiltonian—such as that of the uracil cation with Morse and quartic potentials—can be Trotterized without digitizing the vibrational continuum, with a per-step cost that scales linearly in the number of modes and only logarithmically in the desired accuracy. The authors validate the gate synthesis numerically on the uracil cation and compare resource counts with fully digital encodings, arguing that the main price is a postselection success probability that decays exponentially with the number of anharmonic gates.","feed_headline":"Oscillator-qubit compiler makes anharmonic gates in O(log 1/eps) depth","feed_subtitle":"Hybrid processor simulates vibronic dynamics without discretizing vibrations; speed-up relies on smooth periodic extension.","key_machinery":"The core object is the OQ-GQSP sequence built from two complementary signal operators A_Q = diag(e^{iπQ/L}, I) and B_Q = diag(I, e^{-iπQ/L}), each constructed from two Z-basis conditional-displacement gates and a reusable ancilla. By the GQSP theorem, an alternating product of these operators and qubit rotations realizes a block matrix whose top-left entry is an arbitrary Laurent polynomial F(e^{iπQ/L}) in the displacement operator. This Laurent polynomial is used to approximate the Fourier series of exp(iΔt V(Q)), so the same circuit performs the nonlinear phase gate; a second OQ-GQSP block, followed by qubit measurements, makes the gate state-dependent (apply to one electronic state only).","core_discovery":"The paper's central claim is that arbitrary bosonic phase functions can be generated on an oscillator-qubit processor by a sequence of Z-basis conditional-displacement gates, removing the parity constraints of earlier hybrid QSP and thereby enabling state-dependent nonlinear bosonic phase gates through postselection. Theorem 2 asserts that one Trotter step of an anharmonic vibronic Hamiltonian costs O(N M' ln(1/ε) + N^2 M) conditional-displacement gates with success probability (1-δ)^{2M'N}, where N and M are the numbers of electronic states and vibrational modes and M' the number of anharmonic modes. The paper also shows numerically that the OQ-GQSP simulation of uracil cation captures the","pith_inferences":["The O(log 1/ε) resource claim rests on exponential decay of Fourier coefficients of exp(iΔt V(x)) for the 2L-periodic extension; for the Morse potentials in this paper, V(-L) and V(L) differ by several eV, so the periodic extension is discontinuous and the coefficients decay only algebraically. A direct numerical check of the Fourier coefficient decay for the paper's ν26 mode would determine wheth","The reported fidelity of 0.9229 at d=39 (error ~0.08) is far from the ρ=e^π≈23.14 geometric rate predicted under the strip-analytic assumption, suggesting that the numerical convergence observed is consistent with algebraic rather than exponential decay. Testing higher d values would clarify the actual cost.","If the holomorphic-extension assumption fails only at the boundary, a modified approximation—e.g., using a windowed or L-dependent periodization of the potential—might restore exponential convergence and is a natural extension.","The same Fourier/GQSP synthesis could be extended to multivariate potentials once multivariable QSP angle finding becomes practical, which would allow mode–mode couplings rather than only diagonal anharmonicity."],"forward_implications":["If correct, anharmonic molecular dynamics can be simulated on hybrid oscillators with gate count growing linearly with modes and logarithmically with accuracy, avoiding the exponential cost of MCTDH and the discretization overhead of qubit-only encodings.","The same compiler can prepare vibrational states matching anharmonic potentials, since the phase gates approximate arbitrary analytic potentials and can be used in state preparation.","Success probability can be traded against circuit depth by adjusting the Fourier degree, giving a practical trade-off for near-term hardware.","The protocol extends the reach of analog vibronic simulators by adding diagonal anharmonicity (Morse, quartic) while keeping off-diagonal linear couplings native.","The uracil cation case shows anharmonic terms are necessary to reproduce sub-50 fs population relaxation, so the method addresses a physically relevant failure mode of harmonic models."],"fun_headline_variants":["Hybrid GQSP: any bosonic phase, no parity limits","Oscillator-qubit compiler scales linearly with vibrational modes","Simulate anharmonic vibronics without discretizing vibrations","O(log 1/eps) depth for anharmonic gates on hybrid processors","Uracil cation vibronics made tractable on hybrid processors"],"cache_read_input_tokens":21376,"weakest_assumption_plain":"The claim that the circuit depth grows only as log(1/ε) assumes that each phase function exp(iΔt V(Q)) extends to a holomorphic function on a strip of width σ around the real axis, and for the Morse potentials used in the uracil cation model this periodic-extension analyticity does not appear to hold, since the potential differs by several eV across the period boundary.","fun_headline_variants_meta":{"raw":{"variants":["Hybrid GQSP: any bosonic phase, no parity limits","Oscillator-qubit compiler scales linearly with vibrational modes","Simulate anharmonic vibronics without discretizing vibrations","O(log 1/eps) depth for anharmonic gates on hybrid processors","Uracil cation vibronics made tractable on hybrid processors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001669,"raw_usage":{"total_tokens":6454,"prompt_tokens":735,"completion_tokens":5719,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":479,"completion_tokens_details":{"reasoning_tokens":5632}},"tokens_in":479,"tokens_out":5719,"duration_ms":36565,"temperature":1.0,"reasoning_tokens":5632,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T10:20:13.448197+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Fourier coefficients a_k of g(x)=exp(iΔt V_Morse(x)) on [-L,L] for the paper's ν26 mode parameters and check whether |a_k| decays as C ρ^{-|k|} with ρ=e^{πσ/L} or only polynomially. If the decay is algebraic, the d=O(log 1/ε) theorem does not apply to this case and the observed fidelity 0.9229 at d=39 would be the expected behavior, not an anomaly.","supporting_citations":[],"review_version":1}