{"id":"4e3aec28-fb02-4bf8-8f5a-e16ba5ea5975","arxiv_id":"2505.16286","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A superconducting quantum processor is programmed to realize tunable XYZ, transverse-field Ising, and Dzyaloshinskii-Moriya spin Hamiltonians using single-qubit rotations and native XY coupling.","lead":"Researchers used microwave pulses and the natural coupling between superconducting qubits to create several different types of spin interactions, including anisotropic XYZ, transverse-field Ising, and Dzyaloshinskii-Moriya couplings. The work expands the kinds of quantum spin models that can be studied on superconducting processors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The DM validation is undermined: |ψ0>=V|ψx> is not a zero-energy eigenstate of Eq. (9), so the observed freezing does not confirm the engineered G/D ratio.","rationale":"The reader's weakest assumption about first-order Floquet validity is legitimate but not the most load-bearing issue. The paper never reports τ_i or T_c, and finite-width pulses are not included, yet the XYZ frequency-scaling check provides partial quantitative support, so that concern alone would not change the conditional verdict. A more specific and serious problem is the DM validation protocol. The state |ψ0> = V(G/D)|ψx> is claimed to be a zero-energy eigenstate of Eq. (9), but H_XY+DM conserves total Sz. A product state with transverse magnetization has support on all Sz sectors, and the one-magnon components acquire energies ±|G+iD|, not zero. Therefore |ψ0> is not an eigenstate for any G/D. The observed slow decay of ⟨Sx⟩ is expected for any x-y-plane product state because d⟨Sx⟩/dt vanishes at t=0 when ⟨σz⟩=0; it does not specifically validate the engineered G/D ratio. This does not prove the engineered Hamiltonian is wrong—the numerical curves might still match if they use the actual state—but it means the stated evidence for the DM result is invalid. The verdict should remain CONDITIONAL because the central claim is plausible and the XYZ part is quantitatively supported, but the DM demonstration needs a corrected validation, e.g., direct Hamiltonian tomography or a quantitative fidelity metric. The concrete test above would settle whether the eigenstate claim fails; if it does, the paper must be revised before acceptance. The reader's conditional verdict is therefore retained, but for a different, more specific reason than the Floquet timings.","tokens_in":15944,"tokens_out":20908,"duration_ms":157027,"concrete_test":"Compute H_XY+DM |ψ0> for L=8 using Eq. (9) with the stated G/D values and φ = tan^{-1}(G/D)+π. If |ψ0> is a true zero-energy eigenstate, the norm should be zero. Also simulate the exact noiseless time evolution of ⟨Sx(t)⟩ starting from |ψ0> over 1 μs; if it is an eigenstate, ⟨Sx(t)⟩ is constant. Any oscillation with amplitude exceeding the decoherence scale refutes the eigenstate claim and invalidates the present DM validation. A second check: measure ⟨Sx(t)⟩ for the fully polarized state |0...0> as a control; if it remains constant while the spiral state decays, the difference is due to particle-number sector mixing, not Hamiltonian accuracy.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim rests on three demonstrations. The XYZ part is supported by a quantitative frequency check, and the transverse-field Ising part is standard. The DM interaction, however, is validated by preparing |ψ0> = V(G/D)|ψx> and asserting it is a zero-energy eigenstate of Eq. (9). This assertion is incorrect. H_XY+DM conserves total Sz, so its product zero-energy eigenstates are only the fully polarized states |0...0> and |1...1>. The state |ψ0> is a product state with all spins in the x-y plane; it has support on every Sz sector, and in the one-magnon sector the phase differences (φ = tan^{-1}(G/D)+π) give eigenvalues ±√(G^2+D^2), not zero. For G/D=1, the phase difference per site is 5π/4, whereas the Hamiltonian's hopping phase is π/4; the prepared state is a superposition of eigenstates from different particle-number sectors with different energies. Consequently, the observed 'striking freezing' of ⟨Sx(t)⟩ does not demonstrate stationarity under the target Hamiltonian. Rather, any product state with ⟨σz⟩=0 has zero initial slope d⟨Sx⟩/dt, so the plateau is an artifact of the initial state, not a fidelity probe of the engineered G/D ratio. The numerical simulations shown as solid curves might still validate the Hamiltonian if they use the actual prepared state, but the paper's stated eigenstate protocol does not.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports experimental Floquet engineering of tunable spin interactions on a superconducting transmon processor with tunable couplers. Using a sequence of global pi/2 pulses and resonant XY evolutions, the authors claim to realize (i) XYZ/XXZ spin models with continuously adjustable anisotropy, (ii) transverse-field Ising interactions, and (iii) XY plus Dzyaloshinskii-Moriya (DM) interactions with tunable G/D ratio. The XYZ section includes a quantitative check: the measured frequency reduction factor 0.65(2) agrees with the predicted 2/3, and the anisotropy dependence of Jx-Jz is shown. The transverse-field Ising data are presented with qualitative agreement to numerical simulations. The DM section validates the engineered G/D ratio by preparing a spin-spiral product state claimed to be a zero-energy eigenstate of Eq. (9) and observing suppressed time evolution of the rotated magnetization.","tokens_in":16259,"tokens_out":13491,"duration_ms":114593,"significance":"If the claims hold, the work provides a useful experimental toolbox for programmable spin-Hamiltonian simulation in superconducting circuits, complementing similar Floquet-engineering demonstrations in Rydberg arrays and trapped ions. The XYZ part is supported by a clean quantitative frequency measurement, and the phase-calibration procedures are carefully described. The DM part, however, is the weakest: the validation relies on an eigenstate-stationarity argument that is mathematically incorrect, and no quantitative fidelity metric or error bar is given. Because the central claim of tunable DM interactions is not convincingly established, the paper in its current form does not fully support its abstract. The results are nevertheless likely repairable by re-analyzing the DM data with a correct validation protocol, which warrants a major revision rather than rejection.","major_comments":[{"comment":"The statement that |ψ0⟩ = V(G/D)|ψx⟩ is a zero-energy eigenstate of Eq. (9) is incorrect. The Hamiltonian in Eq. (9) conserves total S_z, so its zero-energy eigenstates in product form are only the fully polarized states |0...0⟩ and |1...1⟩. The spiral product state |ψ0⟩ has support on all S_z sectors, and in the one-magnon sector the eigenvalues are ±(G^2+D^2)^{1/2} (times the appropriate normalization), not zero. Therefore the observed freezing of ⟨S̃x(t)⟩ does not demonstrate stationarity under the engineered Hamiltonian. For any product state with ⟨σz_i⟩=0, the initial slope d⟨Sx⟩/dt vanishes by symmetry, so the early-time plateau is not a fidelity probe of the G/D ratio. The authors should either remove the eigenstate claim and validate the G/D ratio by comparing the full time evolution with numerical simulations starting from the actual prepared state, using a quantitative fidelity measure, or prepare a genuine eigenstate of Eq. (9). As written, the DM interaction demonstration lacks a valid quantitative validation.","section":"DM validation, Fig. 4(d)-(e) and surrounding text"},{"comment":"The effective-Hamiltonian approximation H_eff = (1/Tc) Σ_i τ_i H_i is invoked without reporting the actual time allocations τ_i and total period Tc, nor the 20-ns gate width relative to these intervals. The validity condition J_ij Tc << 2π is stated, but no numerical values are provided. The finite width of the microwave pulses is treated as instantaneous, and no estimate of the resulting correction to the engineered Jx, Jy, Jz is given. Without these numbers, the reader cannot assess whether the quantitative agreement in Fig. 2(c) is consistent with the stated approximation or is partly accidental. Please report the experimental values of τ1, τ2, τ3, Tc, and the gate length, and provide a numerical estimate of the leading Floquet corrections (e.g., via exact Floquet simulation including finite-width pulses) for the parameters used.","section":"Eq. (6) and the XYZ/Ising sequences"}],"minor_comments":[{"comment":"The total period is defined as Tc = 2(τ1+τ2+τ3) in the supplementary material but as Tc = Σ_i τ_i in the main text after Eq. (6). The mapping formulas for Jx, Jy, Jz are consistent once this notational difference is accounted for, but the conflicting definitions will confuse readers and should be harmonized.","section":"Supplementary Material, Eq. (1)"},{"comment":"The transverse-field Ising demonstration is only qualitative: no extracted B or J_t values, no decay rates, and no quantitative comparison metric are given. The error bars are reported as the standard error over three repetitions, but the agreement with simulation is assessed by eye. A quantitative analysis (e.g., fitted oscillation frequency versus B) would strengthen this part of the claim.","section":"Fig. 3(b) and transverse-field Ising section"},{"comment":"The numerical simulations in Figs. 2 and 4 use T1 ~ 10 μs and T2* ~ 1.5 μs, whereas the device table in the supplementary lists T2* values ranging from 0.84 μs to 14.92 μs. The choice of these representative values and whether they were measured at the operating point should be clarified.","section":"Simulation parameters in Figs. 2 and 4"},{"comment":"The sentence 'As illustrated in Fig. 2(a), we implement local Rz modulation...' refers to the DM experiment, but Fig. 2 concerns the XYZ interaction; this should be Fig. 4(a).","section":"Text near Eq. (10)"},{"comment":"The phrase 'high precision preciseness' is a typo and should be corrected.","section":"Summary paragraph"}],"recommendation":"major_revision","confidential_remarks":"The eigenstate validation error in the DM section is a substantive technical flaw, but it appears fixable within the scope of the paper by recasting the validation as a dynamical comparison against exact numerical simulation with the actual prepared state, or by finding a genuinely stationary eigenstate. The XYZ portion is solid and provides a useful quantitative demonstration. The paper is better suited for a rapid-communication style journal; the novelty is incremental relative to existing Floquet-engineering literature, but the experimental execution is clean. I would encourage the editor to request a revision addressing the DM validation and the missing Floquet parameters before considering publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: worth a serious referee, but the DM validation is wrong and needs a rework. The XYZ and transverse-field Ising parts are the meat; the XYZ part is genuinely quantitative and correct.\n\nWhat is actually new: the paper transplants the NMR/Rydberg Floquet transformation toolbox (refs. 27-32) onto a superconducting processor and demonstrates three interaction families with a single control framework. The phase-calibration protocol between microwave fields is careful and, as far as I can tell, correct. The quantitative check of the XYZ Hamiltonian—frequency reduction 0.65(2) versus predicted 2/3—is the strongest evidence and deserves credit. The transverse-field Ising data are qualitative but consistent.\n\nThe soft spot is the DM section. The stress-test note lands. The state |ψ0> = V(G/D)|ψx> is not a zero-energy eigenstate of Eq. (9). H_XY+DM conserves total Sz, so product eigenstates are only fully polarized along z; the spiral product state has support on many Sz sectors and evolves. The observed freezing of ⟨Sx(t)⟩ is largely an initial-slope artifact: for any product state with ⟨σz⟩=0, d⟨Sx⟩/dt vanishes at t=0. So the eigenstate-stationarity argument does not confirm the engineered G/D ratio. The numerical curves in Figs. 4(d,e) may well be honest simulations of the actual prepared state, but the paper's stated protocol is not a valid fidelity probe.\n\nTwo smaller gaps: exact pulse timings and T_c are never reported, and the high-frequency condition J_ij T_c << 2π is invoked but not quantified. These are reporting gaps, not fatal flaws. The DM data also lack error bars; that is a real weakness for a quantitative claim about a tunable ratio.\n\nWho is this for: experimentalists working on analog/digital quantum simulation with superconducting qubits, and theorists wanting a reference on what is achievable with microwave-only control. It is a useful engineering paper, not a conceptual breakthrough. I would send it to review, with the expectation that the DM section be revised to either provide a correct validation (e.g., compare full dynamics against numerical simulation with the actual state) or drop the eigenstate language. If that is done, it is a solid contribution.\n\nRecommendation: engage with it; accept for peer review with major revision.","headline":"Useful toolbox paper with a solid XYZ demonstration, but the DM validation rests on a false eigenstate claim.","tokens_in":16869,"tokens_out":8543,"would_cite":true,"duration_ms":66564,"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":"This paper reports experimentally realized, continuously tunable XYZ, transverse-field Ising, and Dzyaloshinskii-Moriya spin Hamiltonians on a superconducting qubit processor using only microwave pulses and its native XY coupling.","keywords":["quantum simulation","superconducting qubits","Floquet engineering","XYZ spin model","transverse-field Ising model","Dzyaloshinskii-Moriya interaction","microwave control","tunable couplers"],"falsifier":"Repeat the two-qubit XYZ experiment at a fixed anisotropy ratio while increasing the total period $T_c$ by stretching the resonance intervals, and extract the normalized frequency $(J_x-J_z)/J$ from the y-magnetization oscillations; if the first-order average Hamiltonian is valid, this quantity should stay at the value predicted by $\\eta=J_z/J_x$, whereas higher-order Floquet corrections would make it drift systematically with $T_c$. Since the paper reports the zero-order prediction but no sequence timings, this measurement would settle whether the engineered couplings match the claimed values.","tokens_in":15770,"feed_emoji":"⚛️","tokens_out":11874,"duration_ms":92430,"temperature":0.7,"pith_summary":"This paper reports a set of experiments on a superconducting qubit processor that aim to show that one fixed hardware interaction can be reshaped into several different spin Hamiltonians by microwave control alone. The native coupling between neighbouring qubits is of XY form, i.e. an exchange of excitations, and the paper's claim is that by interleaving this coupling with single-qubit rotations and virtual phase gates, the same device can emulate (i) XYZ models with continuously adjustable couplings, (ii) transverse-field Ising models, and (iii) Dzyaloshinskii-Moriya interactions, an antisymmetric exchange that favours non-collinear spins. The demonstrations use two qubits for the first two families and an eight-qubit ring for the third. If the claim is right, a single analogue-digital simulator gains access to a broader class of spin physics without any change in wiring or coupler hardware.","feed_headline":"Microwave pulses turn one native coupling into three spin models","feed_subtitle":"Superconducting qubits emulate XYZ, Ising, and chiral DM interactions using only pulse timing and phases.","key_machinery":"The load-bearing object is the first-order Floquet average-Hamiltonian identity $\\hat H_{\\rm eff}=\\frac{1}{T_c}\\sum_i\\tau_i\\hat H_i$, valid in the high-frequency regime $J_{ij}T_c\\ll 2\\pi$, together with two conjugation rules for the XY interaction. A global $\\pi/2$ rotation about the x- or y-axis converts XY into XZ or YZ, and a local longitudinal rotation with phase difference $\\Delta\\varphi_{ij}$ converts XY into XY plus a Dzyaloshinskii-Moriya (antisymmetric) term. These rules allow the time allocations $\\tau_i$ and phase differences $\\Delta\\varphi_{ij}$ to become the tunable parameters that set the effective couplings.","core_discovery":"On the paper's own terms, the central discovery is a prescription for translating the native XY interaction $\\hat H_{ij}^{XY}=\\frac{1}{2}J_{ij}(\\sigma_i^x\\sigma_j^x+\\sigma_i^y\\sigma_j^y)$ into a programmable set of effective spin Hamiltonians. The prescription is a first-order Floquet average: a periodic sequence of duration $T_c$ is split into intervals $\\tau_i$ in which the system evolves under different rotated versions of the XY interaction, and the dynamics are governed by $\\hat H_{\\rm eff}=\\frac{1}{T_c}\\sum_i\\tau_i\\hat H_i$. Global $\\pi/2$ rotations transform XY into XZ or YZ interactions, so the two-qubit sequence $\\{-X/2,\\,X/2,\\,Y/2,\\,-Y/2\\}$ with three resonance intervals produces an XYZ Hamiltonian whose coefficients $J_x,J_y,J_z$ are linear combinations of the time allocations. The transverse-field Ising model is obtained by splitting the XY evolution into two intervals separated by a $\\pi$ pulse, which cancels the YY terms while keeping XX, and by implementing the field as a virtual Z rotation. The Dzyaloshinskii-Moriya term is obtained from local phase differences: with a phase $\\Delta\\varphi_{ij}$ between neighbouring sites, the XY coupling becomes $G(\\sigma^x_i\\sigma^x_j+\\sigma^y_i\\sigma^y_j)+D(\\sigma^x_i\\sigma^y_j-\\sigma^y_i\\sigma^x_j)$ with $G/D=\\cot\\Delta\\varphi_{ij}$. The eight-qubit ring realizes this for $G/D=0$ and $G/D=1$ and validates the ratio by preparing zero-energy spin-spiral eigenstates and observing their stationarity.","pith_inferences":["Editorial inference: the same Floquet construction should extend to larger and two-dimensional lattices, but the phase-matching constraint $N\\cdot\\Delta\\varphi=2\\pi n$ ties the achievable $G/D$ ratios to the system size, so arbitrary ratios on a fixed lattice would require site-dependent phases or additional interaction blocks.","Editorial inference: the paper never quotes $\\tau_i$ or $T_c$; a natural quantitative follow-up is a study of how the extracted $J_x,J_y,J_z$ drift as $T_c$ grows, which would map the validity boundary of the first-order Floquet formula for this hardware.","Editorial inference: the zero-energy eigenstate stationarity probe used for the DM interaction could be reused as a calibration routine for the relative microwave phase between any pair of qubits, since the decay rate of a deliberately non-eigenstate spiral is sensitive to the realized $G/D$ ratio.","Editorial inference: for two qubits the commutativity of XZ, XY, and YZ components makes the XYZ realization exact in the average-Hamiltonian limit; on more than two qubits these components do not commute, so the same sequence would incur a higher-order Floquet error that grows with the number of qubits, a limitation the two-qubit demonstration does not expose."],"forward_implications":["A single processor can interpolate continuously between XXX, XXZ, and fully anisotropic XYZ dynamics on two qubits, with the anisotropy $\\eta=J_z/J_x$ set by the ratio of time intervals and the magnetization dynamics reflecting the preserved or broken symmetries.","Transverse-field Ising dynamics can be generated with the field strength $B$ controlled by the virtual Z phase, giving a Floquet route to Ising physics that needs no external magnetic field.","XY plus Dzyaloshinskii-Moriya interactions with $G/D=\\cot\\Delta\\varphi$ can be synthesized on any ring whose size satisfies the periodic-boundary condition $N\\cdot\\Delta\\varphi=2\\pi n$; the realized $G/D=0$ and $G/D=1$ cases are the first two members of this family.","Because the required operations are single-qubit gates, virtual Z phases, and the pre-existing XY coupling, the scheme ports to existing tunable-coupler processors without parametric modulation.","In the two-spin XYZ block the rotated components commute, so the target dynamics are realized exactly within one control period rather than only in a Trotterized limit."],"supporting_citations":[{"why":"supplies the first-order Floquet average-Hamiltonian formula (Eq. 6) and the local Rz modulation prescription for spin-exchange models.","marker":"31"},{"why":"demonstrates microwave-engineered programmable XXZ Hamiltonians in Rydberg arrays, the blueprint this paper adapts to superconducting circuits.","marker":"30"},{"why":"provides the Floquet Hamiltonian engineering approach for many-body spin systems that justifies the periodic-driving scheme.","marker":"29"},{"why":"introduces virtual Z gates, which implement the local phases and transverse-field terms without physical pulses.","marker":"33"},{"why":"describes the tunable-coupling scheme that supplies the native XY interaction used throughout.","marker":"22"},{"why":"defines the Dzyaloshinskii-Moriya antisymmetric exchange interaction that the eight-qubit experiment targets.","marker":"25"},{"why":"supplies the microscopic theory of anisotropic superexchange that motivates the DM term's physical content.","marker":"26"}],"fun_headline_variants":["One XY coupling becomes three spin models via microwaves","Microwave pulses convert XY into XYZ, Ising, and chiral","Tunable spin interactions from a single native coupling","Microwave sequences yield XYZ, Ising, and chiral spin models","Programmable spin Hamiltonians from one XY interaction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole construction hinges on the first-order Floquet average $\\hat H_{\\rm eff}=\\frac{1}{T_c}\\sum_i\\tau_i\\hat H_i$ being accurate, which requires the coupling strength times the sequence period to be well below $2\\pi$ and the 20-ns microwave pulses to behave as instantaneous rotations; the paper never reports the actual interval lengths or sequence period, so the size of the finite-pulse and higher-order corrections is left unquantified.","fun_headline_variants_meta":{"raw":{"variants":["One XY coupling becomes three spin models via microwaves","Microwave pulses convert XY into XYZ, Ising, and chiral","Tunable spin interactions from a single native coupling","Microwave sequences yield XYZ, Ising, and chiral spin models","Programmable spin Hamiltonians from one XY interaction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000223,"raw_usage":{"total_tokens":1502,"prompt_tokens":1032,"completion_tokens":470,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":648,"completion_tokens_details":{"reasoning_tokens":389}},"tokens_in":648,"tokens_out":470,"duration_ms":4418,"temperature":1.0,"reasoning_tokens":389,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T15:04:31.471126+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the two-qubit XYZ experiment at a fixed anisotropy ratio while increasing the total period $T_c$ by stretching the resonance intervals, and extract the normalized frequency $(J_x-J_z)/J$ from the y-magnetization oscillations; if the first-order average Hamiltonian is valid, this quantity should stay at the value predicted by $\\eta=J_z/J_x$, whereas higher-order Floquet corrections would make it drift systematically with $T_c$. Since the paper reports the zero-order prediction but no sequence timings, this measurement would settle whether the engineered couplings match the claimed values.","supporting_citations":[{"cited_title":"Nishad , author A","cited_arxiv_id":null,"evidence_quote":"supplies the first-order Floquet average-Hamiltonian formula (Eq. 6) and the local Rz modulation prescription for spin-exchange models."}],"review_version":1}