{"id":"88276dc0-b869-4864-89fb-f7b8e176e0b3","arxiv_id":"2411.09882","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A fast Rydberg blockade SWAP gate can be implemented with continuously modulated driving, with simulated gate errors below 10^-4.","lead":"This paper designs fast SWAP gates for neutral-atom quantum computers by using smoothly varying laser pulses and the Rydberg blockade effect. It matters because a direct, high-fidelity SWAP gate could improve connectivity and speed in large atom arrays without moving atoms.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Finite-blockade waveforms are optimized for a resonant Förster channel with δq=0; the paper never tests or maps them to the standard V|rr><rr| shift, so the 1e-4 existence claim is conditional on that interaction model.","rationale":"I read the paper in good faith. The numerical examples are concrete and the Fourier-coefficient parameterization is reproducible, so the existence of a SWAP-capable waveform within the Förster model is credible. The weak point is not the mathematics but the physical transferability: the finite-blockade claim is the part that makes the protocol experimentally relevant, and it is demonstrated only for a single resonant Förster channel. The standard Rydberg blockade in most experiments is described by an energy shift, and the paper neither derives Eq. (2) from a specific atomic system nor shows that the optimized pulses survive under a shift-type interaction. The ideal B→∞ limit is fine, but the finite-B adaptation is central to the abstract's claim of compatibility with current techniques. I therefore agree with the reader's conditional verdict and would not change it: the concern is significant enough to keep the verdict conditional, but not to reject, because the model is physically legitimate and the waveforms are specified fully enough for independent simulation.","tokens_in":10578,"tokens_out":13935,"duration_ms":154545,"concrete_test":"Take the finite-blockade waveform coefficients in Appendix A (B=2π×125 MHz, δq=0; and Fig. 8 coefficients with B=2π×100 MHz) and numerically propagate the same initial states using H_int = V|rr><rr| with V=2π×125 MHz (and V=2π×1 GHz, plus a range around these values). Compute the SWAP gate error using the Pedersen et al. criterion. If any error exceeds 1e-4, the claim that the protocol adapts to finite Rydberg blockade strengths is not established for the standard shift model. Optionally, re-run the optimization under H_int=V|rr><rr|; if no waveform with error below 1e-4 is found, the existence claim is model-dependent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central existence claim rests on numerical solutions of the two-atom Förster Hamiltonian in Eq. (2), H_q = B|rr><qq'| + B|qq'><rr| + δq|qq'><qq'|. The finite-blockade examples (Sec. IV, Fig. 8, Appendix A) fix δq=0 and B=2π×125 MHz or 2π×100 MHz. This is a resonant two-level Förster model, not the usual blockade shift V|rr><rr|. For far-off-resonant Förster coupling the shift model emerges with V=B^2/δq, but δq=0 has no such limit, so the two models are not interchangeable. The paper asserts that the gates 'adapt to finite Rydberg blockade strengths' but only demonstrates this within Eq. (2), and it never identifies a specific atomic species, Rydberg levels, electric field, or interatomic distance where B=2π×125 MHz and δq=0 are simultaneously available. If an experiment realizes the interaction as a van der Waals shift, or with multiple Förster channels, the same waveforms may not give error below 1e-4. This is a correctness risk for the headline claim, not a consensus disagreement.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a numerical construction of a two-qubit Rydberg-blockade SWAP gate driven by continuously modulated lasers. The authors consider a two-atom model whose single-atom Hamiltonians couple |0> and |1> to |r> and whose two-body interaction is a Förster-resonance term Hq = B|rr><qq'| + B|qq'><rr| + delta_q |qq'><qq'|. They decompose the computational basis into singlet and triplet sectors and optimize Fourier coefficients for the Rabi frequencies and detunings so that the two sectors acquire the correct relative population and phase behavior. Several waveform families are presented: hybrid amplitude-and-frequency modulation, amplitude-only off-resonant driving, resonant amplitude-only driving, and finite-blockade versions with B = 2 pi x 125 MHz and B = 2 pi x 100 MHz at delta_q = 0. Numerical simulations show gate errors below 1e-4, and robustness scans against Rabi-frequency, detuning, and interaction-strength variations are included. The paper claims this establishes the existence of a fast Rydberg-blockade SWAP gate compatible with current experimental techniques.","tokens_in":10876,"tokens_out":7161,"duration_ms":75965,"significance":"If the construction is valid, it would provide a direct two-qubit SWAP gate for neutral-atom platforms, potentially avoiding the usual decomposition into three Controlled-PHASE gates and extending the recent success of synthetic continuously modulated driving to gates with net population transfer. The paper has several concrete strengths: the Hamiltonian and waveforms are explicit; the singlet/triplet decomposition is analytically transparent; the Fourier coefficients are given in full, which would allow independent reproduction; and the robustness scans in Figs. 4 and 5 add useful information about sensitivity to control errors. The main caveat is that the central existence claim is verified numerically in the same closed-system model used to optimize the waveforms, and the finite-blockade examples rely on a specific Förster interaction with delta_q = 0 that is not mapped to the standard Rydberg blockade shift or to a specific experimental configuration. These issues are load-bearing for the headline 1e-4 error claim and need to be addressed before the result can be considered established.","major_comments":[{"comment":"The finite-blockade waveforms are optimized for the resonant Förster model Hq = B|rr><qq'| + B|qq'><rr| + delta_q|qq'><qq'| with delta_q = 0, and the two finite-blockade examples fix B = 2 pi x 125 MHz and B = 2 pi x 100 MHz. This model is not interchangeable with the standard blockade shift V|rr><rr|: the shift model emerges from a far-off-resonant Förster channel with V = B^2/delta_q, which has no delta_q -> 0 limit. Since no atomic species, Rydberg level pair, electric field, or interatomic distance is identified at which B ~ 2 pi x 125 MHz with delta_q = 0 is physically realized, the claim in Sec. I that the gate can adapt to finite Rydberg blockade strengths is demonstrated only within Eq. (2). The authors should either simulate the same waveforms under a V|rr><rr| interaction (with V in the 100 MHz to 1 GHz range) or provide a quantitative mapping to a specific Förster-resonant experimental configuration. Without one of these, the reported 1e-4 error rates for the finite-blockade examples are not a predictive statement about laboratory Rydberg interactions.","section":"Section II, Eq. (2); Section IV, Fig. 5"},{"comment":"The paper reports 'calculated gate errors are less than 1e-4' but never defines the fidelity or error metric in the main text or appendix. Because the Fourier coefficients are obtained by numerical search in exactly the same closed-system Schrödinger model used to compute these errors, the reported errors are in-sample checks. Please state the fidelity measure explicitly (for example, the Pedersen–Mølmer gate fidelity or the average fidelity over the logical subspace), specify how the four initial states are weighted, and add an out-of-sample validation such as random logical states or an independent integration method. This is necessary for the reader to assess whether the claimed 1e-4 error is a controlled prediction or a property of the optimization landscape.","section":"Section III; figure captions of Figs. 2, 3, 7, 8"},{"comment":"The robustness scans in Figs. 4 and 5 vary Rabi frequencies, detunings, and the Förster parameter B within the same model used to design the pulses. They do not address the two-photon example of Fig. 6, nor do they include spontaneous emission, which the text itself identifies as a fundamental limit and estimates only as ~0.5 gamma_r T. For the claim that the gates are compatible with current experimental conditions, please add at least one numerical simulation including Rydberg-state decay, or state quantitatively the maximum allowed decay rate gamma_r for a target error of 1e-4. Without this, the practical significance of the 1e-4 figure remains unclear.","section":"Section IV, Figs. 4 and 5"}],"minor_comments":[{"comment":"The title and text contain spacing artifacts such as 'SW AP' and the abstract contains the grammatically incomplete phrase 'bear considerable resistance some major adverse effects'; these should be corrected.","section":"Abstract and title"},{"comment":"The caption contains a duplicated phrase, 'Rydberg Rydberg blockade', which should be fixed.","section":"Fig. 2 caption"},{"comment":"The Fourier representation is defined with a denominator (2N+1) for a reference time tau = 0.25 microseconds, but the figures and coefficient tables do not explicitly state the total gate duration for each example; adding this would make the waveforms fully reproducible.","section":"Section III"},{"comment":"The text states that single-photon waveforms can be translated to two-photon waveforms by adiabatic elimination, but no explicit two-photon Hamiltonian or final parameter set is given for the comparison in Fig. 6; a short derivation or a precise reference would make this translation checkable.","section":"Appendix A, Fig. 6"},{"comment":"The Morris–Shore transform is cited as an interpretive tool for the linkage structure, but it is not actually applied in the text; either use it explicitly or remove the citation to avoid an unsubstantiated reference.","section":"Section II, Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The central numerical construction is plausible and the singlet/triplet analysis is a genuine strength, but the headline existence claim is currently conditional on the resonant Förster model and on in-sample error reporting. The most important request is a direct test of the finite-blockade waveforms under the standard V|rr><rr| blockade shift, or a concrete experimental mapping of B and delta_q. I would also ask the editor to encourage the authors to state their fidelity metric and, if possible, deposit the integration code, since the paper's reproducibility currently rests entirely on the printed Fourier coefficients."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Xin, Tianze, and Sun answer a real question: can you do a direct Rydberg blockade SWAP gate, rather than decompose into three CZ gates? Their answer is yes, and they back it with a clean singlet/triplet linkage analysis plus explicit Fourier waveforms. The separation of the dynamics into a blockade-insensitive singlet and a triplet that carries the Rydberg interaction is exactly the right way to think about this problem, and the waveforms they list are concrete enough that a competent simulation group could reproduce them.\n\nThe construction is new relative to the CZ literature, and the robustness scans against Rabi and detuning error and against finite B are a sensible first pass. The paper is clearly written. Credit is due for that.\n\nThe soft spots are real but manageable. The finite-blockade waveforms are optimized for a resonant Förster channel, H_q = B|rr><qq'| + B|qq'><rr| with δq = 0. That is not the usual blockade shift V|rr><rr|; the δq = 0 limit is a degenerate energy-transfer process, and the paper never shows how to map B and δq onto a specific atomic species, Rydberg level, electric field, or interatomic distance. So the statement that the gates 'adapt to finite Rydberg blockade strengths' is demonstrated only inside Eq. (2), not against the standard van der Waals blockade. The stress-test note is right about this. It is a gap, not a refutation, but it makes the 1e-4 errors conditional on an interaction model the paper never justifies.\n\nSecond, the reported gate errors are in-sample: the same model used to fit the Fourier coefficients is used to evaluate the gate. The robustness scans help, but there is no independent test, no code or data, and no comparison against the three-CZ decomposition baseline. These are the kinds of things a referee would ask for, not reasons to desk-reject.\n\nWho is this for? Groups working on neutral-atom connectivity and buffer-atom schemes, who want a numeric starting point for a direct SWAP waveform. It is not a finished experimental recipe, but it is a credible existence proof with a clean theoretical core. I would send it to peer review; the referee should focus on the interaction-model mapping and on getting the authors to state the fidelity formula explicitly and provide the waveforms in machine-readable form.\n\nIn short: the central idea holds up, the paper is honest, and the missing piece is a clear bridge from the Förster model to realizable Rydberg parameters.","headline":"Plausible direct Rydberg SWAP gate, but the finite-blockade waveforms are conditional on a resonant Förster model never mapped to the usual V|rr><rr| shift.","tokens_in":11422,"tokens_out":4822,"would_cite":true,"duration_ms":41534,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["32.80.Qk","03.67.Lx","42.50.-p","33.80.Rv"],"model":"deepseek-v4-flash","headline":"The paper establishes that a fast two-qubit SWAP gate can be implemented directly on neutral-atom qubits via Rydberg blockade, with numerically optimized modulated driving waveforms yielding gate errors below 10^-4.","keywords":["Rydberg blockade","SWAP gate","neutral atom qubits","synthetic modulated driving","Fourier series waveforms","two-qubit gate","cold atom quantum computing","quantum gate design"],"falsifier":"Simulate the published waveforms under the standard blockade Hamiltonian V|rr><rr| with V = 2π × 125 MHz, instead of the Forster model in Eq. (2), and compare the resulting gate error; if the error rises above $10^{-4}$, the model-dependence of the design is exposed. Equivalently, an experiment measuring two-qubit state transfer fidelity with these pulses on a pair of atoms at the designed interaction strength would settle the matter.","tokens_in":10344,"feed_emoji":"⚛️","tokens_out":5614,"duration_ms":50757,"temperature":0.7,"pith_summary":"This paper establishes that a fast two-qubit SWAP gate can be implemented directly on neutral-atom qubits using the Rydberg blockade effect, without decomposing the operation into three controlled-phase gates. The authors design laser pulse waveforms, expressed as finite Fourier series with synthetic continuous modulation, that drive the ground-Rydberg transitions of both qubits and produce the correct population exchange and phase accumulation. Numerical simulations of these waveforms show gate errors below $10^{-4}$ under idealized conditions, and the designs adapt to finite Rydberg blockade strengths and tolerate typical laser fluctuations. If the results transfer to the lab, a SWAP gate becomes a basic primitive for cold-atom quantum processors, complementing the existing controlled-phase gates and enabling higher connectivity in large qubit arrays.","feed_headline":"Rydberg blockade SWAP gate is real, error below 1e-4","feed_subtitle":"Direct two-qubit swap is possible with modulated laser pulses, no three-gate decomposition needed.","key_machinery":"The key machinery is the separation of the two-qubit dynamics into singlet and triplet subspaces, defined by the symmetric and antisymmetric combinations of |01> and |10>. Only the triplet couples to the Rydberg blockade through the Forster term B|rr><qq'| + h.c. with penalty δq, while the singlet evolves as a V-type three-level system unaffected by the interaction. The paper encodes the driving fields as truncated Fourier series and optimizes the coefficients numerically to satisfy the SWAP conditions: full population return and the correct relative phase between singlet and triplet. This structure turns the SWAP gate design into an inverse problem solvable by standard optimization, and it is the reason the gate can be made fast and robust.","core_discovery":"The central claim is that the Rydberg blockade SWAP gate protocol does exist, contrary to the implicit assumption that blockade gates are limited to controlled-phase operations. The paper shows that by driving both qubit states |0> and |1> to the same Rydberg level with two coherent lasers, the two-qubit Hilbert space splits into a singlet subspace that evolves without Rydberg interaction and a triplet subspace that couples to the doubly excited state |rr> through a Forster resonance of strength B. The SWAP operation is realized by choosing continuously modulated Rabi frequencies and detunings, found by numerical optimization, so that the singlet acquires a π phase relative to the triplet while all populations return to the computational basis. The paper provides explicit waveform coefficients for the standard SWAP format and an opposite format, using hybrid amplitude-frequency modulation or amplitude-only modulation, for both ideal and finite blockade strengths, and it reports gate errors below $10^{-4}$ in all cases.","pith_inferences":["The same Fourier-series optimization strategy could be applied to design a family of two-qubit exchange gates, not just SWAP, by changing the target phase relation between singlet and triplet.","Combining this SWAP gate with the buffer-atom framework may enable all-to-all connectivity in large arrays, a step toward fault-tolerant neutral-atom computation.","The robustness to laser fluctuations, demonstrated for Rabi frequency and detuning errors, suggests the waveforms could be made even more robust by including those error channels explicitly in the optimization cost function.","The claimed 10^-4 error assumes the Forster resonance model; testing the same waveforms against the standard blockade-shift model would determine how model-dependent the result is."],"forward_implications":["A direct SWAP gate avoids the three-controlled-phase decomposition, shortening the gate sequence and reducing accumulated errors in neutral-atom processors.","The gate operates with interaction strengths on the order of 100 MHz, matching currently available Rydberg blockade setups and enabling fast operation.","The same design method extends to two-photon ground-Rydberg transitions by adiabatic elimination, requiring no extra design burden.","The gate is compatible with the buffer-atom framework for long-range connectivity, allowing distant qubits to swap through buffer atoms without shuttling.","The singlet-triplet decomposition provides a general template for designing other exchange-type gates, such as iSWAP or sqrt(SWAP), via synthetic modulated driving."],"supporting_citations":[{"why":"Supplies the method of suppressing high-frequency components in the driving waveforms, which the SWAP design adopts.","marker":"[34]"},{"why":"Supplies the continuously-modulated driving scheme for controlled-phase gates that the SWAP protocol extends.","marker":"[35]"},{"why":"Supplies the two-atom dark state mechanism that enhances robustness of the gate.","marker":"[37]"},{"why":"Supplies the Morris-Shore transform used to interpret the transition linkage structure.","marker":"[38]"},{"why":"Supplies the buffer-atom framework that the SWAP gate is designed to integrate with.","marker":"[39]"},{"why":"Supplies the earlier numerical pulse optimization approach for Rydberg gates that the waveform search follows.","marker":"[18]"}],"fun_headline_variants":["Direct Rydberg SWAP gate achieves 1e-4 error","Modulated driving enables fast Rydberg SWAP","Rydberg SWAP gate: no decomposition, error 1e-4","Two-qubit SWAP from Rydberg blockade at 1e-4","Synthetic pulses bring Rydberg SWAP to 1e-4"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The designs assume the Rydberg interaction is well described by a Forster resonance coupling |rr> to |qq'> with a given strength B and penalty δq; if the real interaction follows a different model, such as a simple blockade shift V|rr><rr|, the simulated $10^{-4}$ errors may not hold in the lab.","fun_headline_variants_meta":{"raw":{"variants":["Direct Rydberg SWAP gate achieves 1e-4 error","Modulated driving enables fast Rydberg SWAP","Rydberg SWAP gate: no decomposition, error 1e-4","Two-qubit SWAP from Rydberg blockade at 1e-4","Synthetic pulses bring Rydberg SWAP to 1e-4"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000439,"raw_usage":{"total_tokens":2205,"prompt_tokens":897,"completion_tokens":1308,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":1211}},"tokens_in":513,"tokens_out":1308,"duration_ms":9632,"temperature":1.0,"reasoning_tokens":1211,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T20:12:11.745775+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the published waveforms under the standard blockade Hamiltonian V|rr><rr| with V = 2π × 125 MHz, instead of the Forster model in Eq. (2), and compare the resulting gate error; if the error rises above $10^{-4}$, the model-dependence of the design is exposed. Equivalently, an experiment measuring two-qubit state transfer fidelity with these pulses on a pair of atoms at the designed interaction strength would settle the matter.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Morris-Shore transform used to interpret the transition linkage structure."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the earlier numerical pulse optimization approach for Rydberg gates that the waveform search follows."}],"review_version":1}