{"id":"f94f4a87-47df-44f4-b289-cdbc51423691","arxiv_id":"2504.15282","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A single global laser pulse can implement CZZ and CZZZ multi-target gates on one control and two or three target atoms, with simulated fidelities up to 99.55% and 99.24%.","lead":"This paper proposes a way to make one control atom flip several target atoms at once using the Rydberg-blockade effect in neutral-atom quantum computers. If the simulated results hold in real hardware, quantum error correction could run fewer steps and spend less time moving atoms.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Headline fidelities are obtained with V_TT ≈ 0.73–0.78 Ω, so the gate may be compensating for a finite target–target interaction rather than operating in the claimed 'negligible blockade' regime.","rationale":"My primary concern is shared with the reader's weakest_assumption. The central mechanism in §3.1–3.2 is built on V_TT ≪ Ω, and the chosen radii put V_TT at 73–78% of Ω. That does not necessarily make the gate impossible—optimal control can compensate—but it means the paper's advertised physical picture and the 'practical path' claim are not supported by the reported numbers alone. The numbers demonstrate numerical compensation, not the blockade hierarchy. I am not calling the result fraudulent or claiming the gates cannot work; the simulation appears internally consistent. The proposed V_TT = 0 re-evaluation test would settle whether the high fidelity is an artifact of the marginal operating point or a robust consequence of the geometry. Because the reader already conditioned the verdict on this issue, I recommend no change to the verdict; the test should be run before the 'practical path' claim is taken at face value.","tokens_in":16623,"tokens_out":13931,"duration_ms":144602,"concrete_test":"Take the optimized CZZ pulse shown in Fig. 5a (R = 3.5 µm, F = 0.9955) and re-evaluate it in the identical noisy simulation with the target–target term V_TT set to zero, leaving all other parameters and the pulse fixed. If the fidelity stays above ≈99%, the high fidelity does not depend on the residual T–T interaction and the blockade-hierarchy description is acceptable. If the fidelity drops noticeably (e.g., below 98% or by more than the 0.5% error budget), the pulse is actively compensating for a strong T–T interaction, confirming that the operating point violates the 'negligible target-target blockade' assumption in §3.2. Re-optimizing from scratch at V_TT = 0 and comparing the best achievable fidelity would quantify how much of the reported performance is due to that compensation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"For the abstract's central claim to be a statement about asymmetric blockade, §3.2 requires V_TT ≪ Ω. The published operating points violate this: for CZZ at R = 3.5 µm, d_TT = 7.0 µm and V_TT = C6/7^6 ≈ 2π × 7.33 MHz, which is 0.73 of Ω = 2π × 10 MHz; for CZZZ at R = 4.0 µm, d_TT = 2R sin(π/3) ≈ 6.93 µm and V_TT ≈ 2π × 7.80 MHz, which is 0.78 Ω. 'Below Ω' (Fig. 1) is not 'negligible' by the blockade criterion |V| ≫ Ω; at this ratio target–target interactions produce order-unity shifts and partial blockade. The 99.55%/99.24% numbers come from optimizing the full Hamiltonian that contains this term, so they quantify compensation of a strong T–T interaction, not the clean hierarchy claimed in §3.2. The scheme therefore inherits sensitivity to the exact C6 coefficient and to any mechanism (e.g., Rydberg-state choice, tensor effects) that changes V_TT, and the abstract's 'negligible' wording overstates the operating regime.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a scheme for single-control multi-target controlled-Z gates C(Z^⊗N) for N=2 and N=3 on a single-species neutral-atom platform using one global Rydberg pulse. The control atom is placed at the center of a circle and the targets on the perimeter, creating strong control-target blockade and weaker target-target interactions. The authors use a GPU-accelerated pulse optimization with a smoothness penalty to design pulses, reporting fidelities up to 99.55% for CZZ and 99.24% for CZZZ under simulated Rydberg decay and atom-position disorder. They argue that such gates can reduce syndrome-extraction depth in error correction and compare the per-CZ error rate favorably with current two-qubit gates.","tokens_in":16833,"tokens_out":5649,"duration_ms":50052,"significance":"If the numerical results are reliable, the proposed multi-target gates would be a useful primitive for neutral-atom quantum computing, potentially reducing circuit depth for stabilizer readout. The geometric configuration is concrete, the pulse-optimization methodology is standard, and the use of experimentally motivated parameters (Rydberg lifetime, C6 coefficient, trap-position spreads) is a strength. The work includes numerical evidence and a qualitative Pauli-transfer-map analysis of error channels. However, the central physical claim of a 'negligible' target-target interaction is not consistent with the operating parameters, and the reported robustness figures are in-sample optimization results rather than independent predictions, so the significance as stated is overstated.","major_comments":[{"comment":"The paper states that target-target interactions are negligible, but the operating points in Table 2 give V_TT ≈ 2π × 7.33 MHz for CZZ at R = 3.5 μm (d_TT = 7.0 μm) and V_TT ≈ 2π × 7.80 MHz for CZZZ at R = 4.0 μm (d_TT = 6.93 μm), i.e., 0.73–0.78 of Ω = 2π × 10 MHz. This is not negligible under the standard blockade criterion |V| >> Ω; it is a finite interaction comparable to the drive. The high-fidelity gates therefore rely on the optimizer compensating for a significant target-target coupling rather than on the 'negligible' regime described in §3.2. The paper should either (a) choose R such that V_TT << Ω and demonstrate that high fidelity is still achievable, or (b) explicitly reframe the mechanism as 'finite target-target coupling with optimized compensation' and analyze sensitivity to the C6 coefficient and Rydberg-state choice. This is load-bearing because the abstract and conclusion attribute the gate to 'asymmetric blockade' with 'minimal target-target interactions'.","section":"§3.1, Fig. 1, Table 2"},{"comment":"The reported fidelities are in-sample optimization results: the optimizer is trained on the same Gaussian displacement distributions and decay model used for evaluation, and the reported numbers are the average fidelity over these training samples. This does not demonstrate robustness beyond the modeled noise. The paper should state this explicitly and provide an out-of-sample test, e.g., evaluating the optimized pulses on a different noise realization, a different displacement width, or an additional noise channel such as laser intensity noise or Doppler shifts, to quantify generalization. As written, the abstract's claim 'even in the presence of simulated atom placement errors and Rydberg-state decay' is weaker than it appears.","section":"§4.3, Table 1, Table 2"},{"comment":"The 'per-CZ error rate' defined as (1 - F)/N is not a meaningful comparison with the error rate of a single two-qubit CZ gate, because a multi-target gate has a different error structure and a single pulse can fail in ways that affect multiple targets coherently. The claim that the gates have 'error per CZ below the 0.5% level' should be justified by a metric relevant for the actual use case (e.g., logical error rate after syndrome extraction) or by a comparison of the full process fidelity with the equivalent sequence of N two-qubit gates. Without this, the comparison to current two-qubit gates is misleading.","section":"§4.4"}],"minor_comments":[{"comment":"The description of the Kraus map for Rydberg decay is confusing: the text says 'Modeling the |r>→|e> transition comprises omitting the p2 kraus channel', but the decay probabilities p0, p1, p2 are defined and the Kraus operators E0, E1, E2 are given. Please clarify how p2 is used and ensure the map is trace-preserving if p2 is omitted.","section":"§3.3"},{"comment":"The PTM analysis is qualitative; please provide the explicit values of the dominant off-diagonal elements or a quantitative decomposition of the infidelity contributions, so the reader can verify the claim that X↔Y and Z↔I errors dominate.","section":"§4.2, Appendix A"},{"comment":"The y-axis is on a logarithmic scale; please ensure the legend and axis labels are fully readable, and clarify in the caption that the 'Control-Target Interatomic Distance' in the legend refers to the radius R.","section":"Fig. 3"},{"comment":"The companion work [22] is referenced as 'in a companion work'; the claim that this scheme 'enables a depth-optimal readout' is speculative and should be phrased as a potential rather than an established result, or supported by the companion paper.","section":"§5, Ref. [22]"},{"comment":"The paper acknowledges laser intensity and phase uniformity as an experimental concern, but no quantitative analysis is provided; given that the gate relies on precise phase control, a discussion of the acceptable intensity inhomogeneity would strengthen the experimental-feasibility section.","section":"§4.5"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about this paper. First, it is a legitimate numerical proposal for single-pulse C(Z⊗2)/C(Z⊗3) gates in a single-species neutral-atom system, and it may actually be useful for syndrome extraction depth. Second, the 'negligible target-target blockade' framing is overstated: at the operating points they report, V_TT is roughly 0.73–0.78 Ω, so the gate relies on the optimizer compensating for a substantial T-T interaction, not on the clean hierarchy the text describes.\n\nCredit where due: the circular geometry (control at center, targets on perimeter) is a concrete and testable idea; the GPU-accelerated optimizer with smoothness constraints is sensible; the PTM error analysis is useful; and they are upfront about N>3 being out of reach. The paper also cites the relevant prior work on asymmetric blockade (e.g., Young et al.) rather than pretending the concept is new.\n\nSoft spots, in order of concern. The regime issue above is the main one: the abstract says target-target blockade is 'negligible', but the numbers say otherwise. That matters because if the gate is compensating a finite V_TT, its fidelity is sensitive to the exact C6 coefficient and the Rydberg state, which the paper does not address. Second, the reported fidelities are in-sample: same displacement/decay distributions for training and evaluation, so they are optimization metrics, not predictions. There is no held-out set, no code, no experiment. The comparison to ~0.5% experimental per-CZ error is not apples-to-apples; a simulated gate with a specific noise model is not the same as a hardware demonstration.\n\nI would not take the per-target errors below 0.3% at face value, but the central claim—that such pulses exist for this geometry in the model—holds up. The paper is clearly written and honest about its scope. It deserves a serious referee, mostly to push on the regime issue and the in-sample evaluation, and to ask the authors to reframe the abstract and Section 3.2 so 'negligible' matches the actual operating regime. If you work on Rydberg gates or neutral-atom QEC, bring it to your reading group and cite it as a promising proposal—just not as a clean demonstration of asymmetric blockade.","headline":"A solid numerical proposal for single-pulse multi-target Rydberg gates, but the 'negligible target-target' regime is overstated and the headline fidelities are in-sample.","tokens_in":17419,"tokens_out":3020,"would_cite":true,"duration_ms":28348,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.Lx"],"model":"deepseek-v4-flash","headline":"A single global laser pulse on one-species neutral-atom hardware can implement two- and three-target controlled-Z gates, with simulated fidelities up to 99.55% under Rydberg decay and atom-position disorder.","keywords":["neutral-atom quantum computing","Rydberg blockade","multi-target gates","controlled-Z gates","quantum pulse optimization","quantum error correction","syndrome extraction","van der Waals interactions"],"falsifier":"Run the designed CZZ pulse on a single-species neutral-atom array with the control at the center and two targets at $R=3.5$ µm, and measure the process fidelity under the same decay and position-error conditions; the claim is refuted if the per-target infidelity is not below the roughly 0.5% error of two sequential standard CZ gates. A sharper check is to sweep $R$ across the 3.25–3.75 µm band and verify the predicted peak-and-falloff shape in fidelity rather than a flat response.","tokens_in":16372,"feed_emoji":"⚛️","tokens_out":14981,"duration_ms":122781,"temperature":0.7,"pith_summary":"This paper proposes a way to run multi-target controlled-Z gates on a single-species neutral-atom quantum computer without adding extra laser or microwave controls. The key idea is to engineer an asymmetric Rydberg blockade: place the control atom at the center of a circle and the target atoms on its perimeter, so each control–target pair is strongly blockaded while target–target pairs are only weakly interacting. A numerically optimized global laser pulse then implements $C(Z^{\\otimes 2})$ (CZZ) and $C(Z^{\\otimes 3})$ (CZZZ) in a single shot, with simulated fidelities up to 99.55% and 99.24% under Rydberg decay and atom-position disorder. The payoff is that stabilizer syndrome extraction—the repeated measurement step that dominates quantum error correction—could replace a sequence of $N$ two-qubit gates with a single entangling pulse.","feed_headline":"One pulse implements two- or three-target controlled-Z gates","feed_subtitle":"Asymmetric blockade runs two or three CZ gates in one pulse, cutting syndrome-extraction circuit depth.","key_machinery":"The load-bearing object is an asymmetric van der Waals blockade engineered purely by geometry. With $V_{ij}=C_6/d_{ij}^6$, the circular arrangement—control at the center, $N$ targets evenly spaced on the perimeter—makes control–target interactions strong ($V_{CT}\\gg\\Omega$) while keeping target–target interactions weak ($V_{TT}\\ll\\Omega$); this lets the control atom blockade all targets at once while two targets can still be excited together. The asymmetry is what allows one global pulse to attach distinct phases to different multi-qubit states, which a symmetric all-blockaded arrangement cannot do. The second piece of machinery is the pulse-synthesis routine: a gradient-based optimizer with trotterized time evolution, local Kraus amplitude-damping maps for Rydberg decay, mini-batch sampling of random atomic displacements, and a smoothness penalty that pushes the discovered pulses toward experimentally implementable waveforms.","core_discovery":"The central claim is that deliberately breaking the spatial symmetry of the Rydberg blockade turns an all-to-all interaction into a directional one. With the control at the origin and $N$ targets at radius $R$, the control–target interaction is $V_{CT}=C_6/R^6$, while neighboring targets sit at $d_{TT}=2R\\sin(\\pi/N)$, giving $V_{TT}=C_6/d_{TT}^6$. For $R\\approx 3.5$ µm (CZZ) and $R\\approx 4.0$ µm (CZZZ), this places $V_{CT}$ far above the global Rabi frequency $\\Omega$ and $V_{TT}$ near or below it, creating the asymmetry needed for $C(Z^{\\otimes 2})$ and $C(Z^{\\otimes 3})$ phase patterns. The authors synthesize smooth global Rabi pulses by gradient-based optimization that averages over Rydberg decay and Gaussian atom displacements, reaching fidelities up to 99.55% (CZZ) and 99.24% (CZZZ); their error analysis attributes the remaining infidelity mainly to $X\\leftrightarrow Y$ crosstalk on the control qubit and to $Z\\leftrightarrow I$ leakage from residual Rydberg population. The geometry does not extend to larger $N$: once target–target spacing shrinks toward the control–target distance, the required blockade hierarchy breaks down, so the method is presented as a practical primitive for $N=2,3$.","pith_inferences":["If the simulated fidelities survive an experimental test, the per-target error rates would make these gates immediately useful in error-correction circuits, since they dip below the per-CZ error of current two-qubit gates.","The predicted dependence on radius $R$ is sharp enough to be a clean experimental signature: re-optimizing at radii outside the 3.25–3.75 µm band should yield noticeably lower CZZ fidelity, and observing no such falloff would suggest the pulses are compensating rather than using the blockade hierarchy.","The same asymmetric-geometry idea could transfer to other Rydberg-interacting platforms, such as trapped ions or silicon donors, but the pulse shapes would need to be re-derived for each interaction potential and trap geometry.","An analytic reduction of the optimized phase profiles might reveal a closed-form pulse family, which would make calibration simpler and transferable across hardware."],"forward_implications":["A $C(Z^{\\otimes 2})$ or $C(Z^{\\otimes 3})$ gate can be executed in one global pulse lasting hundreds of nanoseconds, so two or three controlled-Z operations take roughly the time of a single standard CZ gate.","The reported fidelities—99.55% for CZZ and 99.24% for CZZZ—correspond to per-target errors of 0.225% and 0.253%, below the roughly 0.5% error reported for current two-qubit gates.","Used in syndrome extraction, these gates cut the number of entangling layers; the companion work shows they enable depth-optimal readout of surface-code and color-code stabilizers.","The circular geometry supports high fidelity only for $N=2$ and $N=3$ targets at current noise levels; $N=4$ or $N=5$ may become feasible with tighter atom positioning, and larger $N$ is excluded by the geometry.","The optimized pulses are smooth and use control hardware already present in state-of-the-art neutral-atom experiments, so no additional laser or microwave fields are required."],"supporting_citations":[{"why":"It supplies the experimental reference point: the Rydberg lifetime, decay branching fractions, gate parameters, and AWG hardware used in the noise model and pulse-implementability checks.","marker":"[12]"},{"why":"It provides the Rydberg-blockade Hamiltonian and the perturbation-theoretic CZ construction from which the optimized pulse dynamics are built.","marker":"[16]"},{"why":"It is the companion work that shows how CZZ and CZZZ gates give depth-optimal stabilizer readout, motivating the target primitives.","marker":"[22]"},{"why":"It supplies the single-species neutral-atom platform assumptions and the $C_6$ blockade coefficient used in the interaction model.","marker":"[26]"},{"why":"They provide the atom-position fluctuation widths used for the displacement noise in the robust pulse training.","marker":"[34, 38]"},{"why":"They supply the Pauli Transfer Map method used to identify the dominant error channels of the optimized gates.","marker":"[39,40]"}],"fun_headline_variants":["One pulse, two or three CZ gates","Multi-target CZ in one shot","Spatial asymmetry enables multi-target CZ","Rydberg geometry yields multi-target CZ"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole scheme rests on the asymmetric-blockade hierarchy being reliable: the control–target interaction must stay much stronger than the Rabi drive while the target–target interaction stays much weaker, and at the chosen radii the target–target interaction is only marginally below the drive, so the optimized pulse must compensate for that residual coupling.","fun_headline_variants_meta":{"raw":{"variants":["One pulse, two or three CZ gates","Multi-target CZ in one shot","Spatial asymmetry enables multi-target CZ","Rydberg geometry yields multi-target CZ"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001938,"raw_usage":{"total_tokens":7617,"prompt_tokens":1018,"completion_tokens":6599,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":634,"completion_tokens_details":{"reasoning_tokens":6542}},"tokens_in":634,"tokens_out":6599,"duration_ms":43161,"temperature":1.0,"reasoning_tokens":6542,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:28:54.982757+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the designed CZZ pulse on a single-species neutral-atom array with the control at the center and two targets at $R=3.5$ µm, and measure the process fidelity under the same decay and position-error conditions; the claim is refuted if the per-target infidelity is not below the roughly 0.5% error of two sequential standard CZ gates. A sharper check is to sweep $R$ across the 3.25–3.75 µm band and verify the predicted peak-and-falloff shape in fidelity rather than a flat response.","supporting_citations":[{"cited_title":"High-fidelity parallel entangling gates on a neutral-atom quantum computer","cited_arxiv_id":null,"evidence_quote":"It supplies the experimental reference point: the Rydberg lifetime, decay branching fractions, gate parameters, and AWG hardware used in the noise model and pulse-implementability checks."},{"cited_title":"Time- optimal two-and three-qubit gates for ryd- berg atoms","cited_arxiv_id":null,"evidence_quote":"It provides the Rydberg-blockade Hamiltonian and the perturbation-theoretic CZ construction from which the optimized pulse dynamics are built."},{"cited_title":"Reduced depth syndrome extraction for neutral atom quantum error correc- tion","cited_arxiv_id":null,"evidence_quote":"It is the companion work that shows how CZZ and CZZZ gates give depth-optimal stabilizer readout, motivating the target primitives."}],"review_version":1}