{"id":"c403c294-0a03-4b1d-b2f8-54b28029302e","arxiv_id":"2602.11576","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Two superconducting resonators shared by two qubits act as a tunable coupler, switching qubit-qubit coupling from zero to above 5 MHz with a ~50 MHz qubit frequency shift.","lead":"Experiments on a two-qubit superconducting chip show that the qubit-qubit coupling can be tuned from fully off to above 5 MHz by shifting qubit frequencies by about 50 MHz, using two shared resonators as a coupler. The result offers a simpler, potentially lower-noise alternative to transmon couplers for large-scale quantum processors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central tuning claim is not independently supported: Eq. (2) omits the gab term present in Eq. (1), is used at moderate g/Δ where perturbative corrections are non-negligible, and the switch-off point is fixed by choosing g12=0.88 MHz as a fit parameter.","rationale":"I agree with the reader that the quantitative claims are conditional. My stress-test adds a more fundamental theoretical concern: the interpretation of the measured anti-crossing gaps relies entirely on Eq. (2), a second-order perturbative expression that (a) drops a term (gab) explicitly retained in the model Hamiltonian and (b) is applied at moderate coupling-to-detuning ratios where higher-order corrections are not guaranteed to be small. Because the two resonator contributions largely cancel at the switch-off point, even a small correction to Eq. (2) can materially alter the inferred coupling. The fact that the theory curve is made to agree with the data by setting g12=0.88 MHz means the data cannot simultaneously be used to validate Eq. (2). This does not undermine the qualitative observation of a tunable anti-crossing minimum, but it does mean the 0-to-5 MHz numbers and the 'switching off' claim are not yet quantitatively secured. An exact-diagonalization cross-check using independent g12 is straightforward and would settle the question. Since the reader's CONDITIONAL verdict already captures the need for such validation, I recommend UNCHANGED.","tokens_in":10448,"tokens_out":7393,"duration_ms":78825,"concrete_test":"Numerically diagonalize Eq. (1) for the stated parameters (ω_a=4.47 GHz, ω_b=4.80 GHz, qubit frequencies around 4.6-4.7 GHz, qubit-resonator couplings g_a1≈g_a2≈27 MHz, g_b1≈g_b2≈30 MHz, anharmonicities from transmon design) over a range of g12 (0-2 MHz) and gab (0-5 MHz). For each qubit-1 frequency, compute the lowest two-qubit-like energy splitting near resonance and compare with Eq. (2)'s 2g_eff and with the six experimental gaps in Fig. 3. If the exact splitting deviates from 2g_eff(Eq.2) by more than ~15% at the operating detunings, or if the switch-off frequency shifts by more than ~10 MHz when g12 is fixed to an independently simulated value rather than fitted, then the central tuning claim requires re-analysis.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline result—that a ~50 MHz shift tunes g_eff from ~0 to >5 MHz—is derived from anti-crossing gaps interpreted as 2g_eff via the perturbative expression Eq. (2). Three features of that derivation are load-bearing. First, Eq. (2) sums only qubit-resonator paths plus g12; it omits the resonator-resonator coupling gab that nevertheless appears in the starting Hamiltonian Eq. (1). If gab is nonzero (no bound is stated), it provides an additional mediation path for qubit-qubit coupling and shifts the switch-off condition. Second, the perturbative formula assumes g_lambda_beta / Delta << 1, but in the relevant operating range the qubit-resonator detunings are only ~100-200 MHz against g~27-30 MHz, i.e., ratios ~0.15-0.3; higher-order terms can shift the cancellation point because the two resonator contributions are designed to nearly cancel. Third, the agreement of Eq. (2) with the measured switch-off point is obtained by taking g12=0.88 MHz as a fitting parameter, not an independent measurement, so the observation of a zero-coupling point is not a test of Eq. (2). Consequently, the quantitative 0-to-5 MHz claim is not established beyond the perturbative approximation; a full treatment might shift the inferred couplings or even the location of the switch-off point.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental study of a double-resonator tunable coupler for two superconducting transmon qubits. The authors measure two-tone spectroscopy anti-crossing gaps and time-domain vacuum-Rabi oscillations as the qubit frequencies are tuned relative to two fixed-frequency resonators. They claim that a roughly 50 MHz shift of the qubit frequencies tunes the effective qubit-qubit coupling from a switching-off point (g_eff ≈ 0) to a two-qubit-gate regime (g_eff > 5 MHz), and that switching off can occur even without a direct qubit-qubit coupling. The supporting theory, Eq. (2), is taken from a prior same-group paper [13], and the direct qubit-qubit coupling g12 = 0.88 MHz is chosen to match the measured switch-off point. The time-domain data are noisy and use a raw readout contrast Δ|IQ| rather than calibrated qubit population; the authors explicitly state that Z-pulse distortion was not calibrated and that the measurements were restricted to small-pulse regimes near 4.637 GHz (Appendix B).","tokens_in":10808,"tokens_out":3375,"duration_ms":37055,"significance":"If the central quantitative claim is established, the paper would provide a useful experimental demonstration of a resonator-based tunable coupler with a potentially small footprint and reduced flux-noise overhead, complementing the more common transmon-coupler architectures. The authors deserve credit for performing both frequency-domain and time-domain measurements across several operating points, and for being transparent about the low SNR, the lack of a Josephson parametric amplifier, and the uncalibrated Z-pulse distortion. However, the headline 0-to-5 MHz tuning range is not yet independently supported: the switch-off point is fixed by fitting g12, the perturbative formula used to convert anti-crossing gaps into g_eff omits the resonator-resonator coupling gab that appears in the starting Hamiltonian, and the time-domain data are qualitative. The experimental observation of a shrinking anti-crossing gap is visible in the data, but the quantitative extraction of g_eff needs a more careful full-Hamiltonian treatment and error analysis before the central claim can be accepted.","major_comments":[{"comment":"Eq. (2) sums qubit-resonator paths plus a direct g12 term, but the starting Hamiltonian Eq. (1) explicitly includes a resonator-resonator coupling gab (c_a† c_b + c_b† c_a ...). If gab is nonzero, it provides an additional mediation path for qubit-qubit interaction and shifts the switch-off condition. The manuscript gives no bound or estimate for gab, so the derived switch-off point may be systematically displaced. This is load-bearing because the central claim that g_eff passes through zero at the observed bias point depends on Eq. (2). Please either measure/place a bound on gab and include its contribution, or justify its omission quantitatively.","section":"Section II, Eq. (1)-(2)"},{"comment":"The quantitative anti-crossing-gap analysis assumes that the measured splitting is exactly 2|g_eff| with g_eff given by Eq. (2). The qubit-resonator couplings are stated as ~27-30 MHz while the relevant detunings in the operating range are only ~100-200 MHz (ratios 0.15-0.3), so the perturbative expression Eq. (2) is used outside its strict g/Δ << 1 regime. Because the two resonator contributions are designed to nearly cancel, relative errors from higher-order terms can be amplified and can move the apparent cancellation point. The paper should compare Eq. (2) with a numerical diagonalization of Eq. (1) (or a dispersive Schrieffer-Wolff calculation keeping higher orders) over the fitted parameter range, and show that the extracted g_eff values remain valid.","section":"Section III, Fig. 3 and text after it"},{"comment":"The direct qubit-qubit coupling g12 = 0.88 MHz is chosen so that the calculated switch-off point coincides with the measured anti-crossing minimum. This makes the agreement a postdiction, not an independent validation of Eq. (2). No independent calibration of g12 (e.g., from a separate two-qubit spectroscopy or electrostatic simulation) is provided, and no error bars are given for the anti-crossing gaps. Statements such as 'reduce to below 2 MHz' and 'almost invisible' are not quantified with statistical uncertainties. Please provide an independent determination of g12 and report uncertainties on the extracted g_eff values.","section":"Section III, 'By choosing direct qubit-qubit coupling as 0.88 MHz...'"},{"comment":"The time-domain evidence is explicitly compromised for quantitative purposes: the authors state in Appendix B that Z-pulse distortion was not calibrated and that the vacuum-Rabi measurements were restricted to small-pulse regimes near 4.637 GHz to avoid distortion. The readout signal is plotted as Δ|IQ| = |IQ| - baseline, not as a calibrated qubit population, and the low SNR (no Josephson parametric amplifier, base temperature above 25 mK) is acknowledged. Therefore the time-domain data can support only a qualitative trend of the envelope changing with flux amplitude; they cannot independently confirm the quantitative 0-to-5 MHz tuning claim or the exact location of the switch-off point. The manuscript should either recalibrate the readout and Z-pulse response or explicitly state that the time-domain data are qualitative only.","section":"Section IV and Appendix B"},{"comment":"The claim in the abstract and conclusions that 'switching off can be realized without direct qubit-qubit coupling' is not demonstrated experimentally. In the experiment g12 is fitted to be 0.88 MHz, i.e., nonzero, and there is no measurement on a device where g12 is engineered to be zero. The statement is a theoretical consequence of Eq. (2), not an experimental result. Please separate the theoretical prediction from the experimental observation, and note that the experiment only shows switching off for one particular nonzero g12 value.","section":"Section III and V"}],"minor_comments":[{"comment":"There are numerous typographical and grammatical errors: 'qubit-qbuit' in Section III, 'Josephosn' in Appendix B, 'respectably' in Fig. 4 caption, and duplicated panel label '(g)' in Fig. 6 caption. A thorough language edit is needed.","section":"Throughout"},{"comment":"The text says 'If qubit-2 is tuned to about 4.37 GHz' in the paragraph describing Fig. 3(e); this appears inconsistent with the figure's frequency ranges near 4.62-4.63 GHz. Please correct the quoted frequency or clarify the typo.","section":"Figure 3 text"},{"comment":"The Rabi-response phase maps in Fig. 8 use color scales in units of 10^-4 (presumably radians) but the axis label says 'Rabi response phase (rad)' without the scaling factor. Please make the units and scaling explicit.","section":"Appendix B, Fig. 8"},{"comment":"The Hamiltonian is written with factors of 1/2 in front of the resonator and qubit terms; the standard notation usually writes ω a†a without 1/2. This is not incorrect if the convention is defined, but please state the convention explicitly to avoid confusion.","section":"Section II, Eq. (1)"},{"comment":"Reference [13] is the same group's earlier theoretical paper from which Eq. (2) is taken. Since the central analysis depends on that formula, the manuscript should cite it prominently in the derivation and clarify exactly which steps are new in this experimental work compared with [13].","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper reports a plausible experimental demonstration of a resonator-based tunable coupler, and the qualitative anti-crossing behavior is visible in the data. My main concern is that the quantitative 50 MHz / 0-to-5 MHz claim rests on Eq. (2) at moderate g/Δ, with g12 adjusted to match the switch-off point and the resonator-resonator coupling gab omitted. These issues are fixable in a revision by adding a full-Hamiltonian analysis, independent g12 calibration, uncertainty estimates, and a more cautious wording of the time-domain and 'without direct coupling' claims. I therefore recommend major revision rather than rejection; the experimental raw data may well support the qualitative conclusion if reanalyzed more carefully."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the main thing you should know: this is the first experimental demonstration of a double-resonator tunable coupler for superconducting qubits. That is the paper's real contribution. The frequency-domain data show anti-crossing gaps between the two qubits shrinking as the qubit frequencies are tuned, with a clear switch-off point near 4.637 GHz. The time-domain vacuum Rabi envelopes follow the same trend, though noisily. The qualitative observation is solid.\n\nWhat is slack: the quantitative claim—that a 50 MHz shift tunes the coupling from ~0 to >5 MHz—is loaded with assumptions. Eq. (2), the effective-coupling formula, comes from the same group's earlier theory paper and omits the resonator-resonator coupling gab that is in the starting Hamiltonian Eq. (1). If gab is finite, it adds another path and shifts the switch-off condition. The formula is also used at g/Delta ~0.15–0.3, where higher-order terms matter for a near-cancellation point. More importantly, the agreement with the measured switch-off point is obtained by choosing g12=0.88 MHz to fit the data; it is not an independent measurement. So the zero point is real, but the exact magnitude of the coupling at each detuning is not nailed down. There are also no error bars on the extracted gaps, and the time-domain data are low-SNR with an explicitly uncalibrated Z-pulse distortion (Appendix B). The authors acknowledge this limitation directly, which is to their credit. The paper also stops short of a two-qubit gate, so 'two-qubit gate point' is an extrapolation.\n\nThe architecture advantages (less flux noise, fewer cables, simpler fabrication) are asserted, not measured. That is fine as motivation, but not as a result.\n\nIn sum: this is a useful, incremental experimental data point for a specific coupler design. It is not a paradigm shift, and the quantitative story needs work. But the central observation—that a double-resonator coupler can have a tunable switch-off point—is visible in the data and worth publishing.\n\nMy recommendation: send it to peer review. Ask for an independent determination of g12, a bound on gab, error bars on the extracted couplings, and, ideally, a two-qubit gate demonstration or at least a calibrated pulse-to-frequency map. The qualitative result deserves to be in the literature; the quantitative claims need to be reined in or properly supported.","headline":"First experiment on a double-resonator tunable coupler: the qualitative switch-off is real, but the 0-to-5 MHz claim rests on a fitted g12 and an uncalibrated Z-pulse response.","tokens_in":11304,"tokens_out":3047,"would_cite":true,"duration_ms":31629,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.Lx","85.25.-j"],"model":"deepseek-v4-flash","headline":"Two fixed superconducting resonators can switch the qubit-qubit coupling from zero to a working two-qubit gate range with only a 50 MHz frequency shift.","keywords":["superconducting qubits","tunable coupler","resonator-mediated coupling","effective qubit-qubit coupling","vacuum Rabi oscillations","Xmon","two-qubit gate","frequency tuning"],"falsifier":"Measure the qubit-qubit coupling at the claimed switch-off point with a calibrated pulse sequence that corrects Z-pulse distortion and independently determine g12; if a nonzero anti-crossing gap or oscillation envelope persists there, or if the coupling at the 50 MHz-shifted point is below 5 MHz, the claimed cancellation would be called into question.","tokens_in":10342,"feed_emoji":"⚛️","tokens_out":6632,"duration_ms":60717,"temperature":0.7,"pith_summary":"This paper reports experiments on a superconducting circuit in which two qubits share two fixed-frequency resonators. The authors show that the effective qubit-qubit coupling is the sum of two opposing resonator-mediated interactions plus a small direct coupling, so tuning the qubit frequency between the resonator frequencies makes the contributions cancel at a switch-off point. In frequency-domain spectroscopy the anti-crossing gap shrinks from about 10 MHz to below the noise floor, and in time-domain vacuum Rabi oscillations the energy-exchange envelope weakens at the same point. Shifting the qubit frequency by roughly 50 MHz from that point restores an effective coupling above 5 MHz, in the range used for two-qubit gates. The claim matters because a simple two-resonator coupler could replace dedicated tunable coupler elements, reducing fabrication complexity and flux-noise sensitivity.","feed_headline":"Qubit coupling switches from off to working gate with a 50 MHz shift","feed_subtitle":"Experiment: two fixed resonators cancel each other's coupling at one detuning and exceed 5 MHz just 50 MHz away.","key_machinery":"The load-bearing object is the effective-coupling formula g_eff = Σ [gλ1gλ2/Δλβ − gλ1gλ2/Σλβ] + g12, which expresses the qubit-qubit interaction as the sum of each resonator's virtual-exchange and counter-rotating contributions plus the direct capacitance between qubits. The paper uses it to predict a cancellation point between the two resonator frequencies and to fit the measured anti-crossing gaps. The experimental probe is the two-tone anti-crossing splitting (2g_eff) in spectroscopy and the vacuum Rabi oscillation envelope in the time domain.","core_discovery":"In the double-resonator coupler circuit, the effective qubit-qubit interaction is described by the sum over the two resonators of (gλ1gλ2/Δλβ − gλ1gλ2/Σλβ) plus a direct qubit-qubit term. When both qubits sit between the two resonator frequencies, the low-frequency resonator contributes a positive interaction and the high-frequency resonator a negative one; at a particular qubit detuning these cancel. The paper reports observing this cancellation directly: the two-qubit anti-crossing gap falls from about 10 MHz to an invisible level as qubit-1 is swept past qubit-2, and the fitted switch-off point agrees with the formula when the direct coupling is taken as 0.88 MHz. Vacuum Rabi oscillation","pith_inferences":["If the cancellation is robust, the residual ZZ coupling at the switch-off point should be directly measurable with a Ramsey/echo experiment; the paper does not report such a direct measurement, but it is a natural next step.","The direct coupling g12 = 0.88 MHz is a fitted value chosen to match the data; an independent extraction of g12, for example from a separate sample with resonators far detuned, would test the formula's predictive power.","With calibrated Z-pulse distortion compensation, the same double-resonator architecture could be operated at larger frequency excursions, extending the gate range beyond 5 MHz.","The cancellation condition generalizes to other qubit types coupled to two resonator modes, suggesting a design principle for modular multi-qubit chips."],"forward_implications":["Two-qubit gates can be turned on with a compact ~50 MHz frequency excursion, so the gate operating point stays close to the sweet spot and flux noise is suppressed.","The qubit-qubit interaction can be switched off completely without any direct qubit-qubit coupling, which also removes static ZZ interactions.","No dedicated flux line for a tunable coupler is needed, reducing cryostat cabling and potential noise sources.","The resonator couplers can be made with narrower coplanar waveguides, shrinking the chip area per qubit in multi-qubit processors.","The same cancellation mechanism offers a way to scale up two-qubit gates without the overhead of transmon coupler tuning lines."],"fun_headline_variants":["Double resonator coupler tunes qubit coupling with 50 MHz shift","Qubit-qubit coupling off-on via 50 MHz frequency shift","Two resonators cancel qubit coupling, then turn on with shift","From zero to 5 MHz: qubit coupling via double resonator","Qubit coupling switch: 50 MHz detuning flips from off to gate"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The reported 0-to-5 MHz tuning range assumes that the anti-crossing splitting is exactly 2g_eff and that the time-domain Δ|IQ| envelope is a faithful vacuum-Rabi signal; the paper notes that Z-pulse distortion was not calibrated, and measurements were restricted to small pulse amplitudes near 4.637 GHz to avoid that distortion.","fun_headline_variants_meta":{"raw":{"variants":["Double resonator coupler tunes qubit coupling with 50 MHz shift","Qubit-qubit coupling off-on via 50 MHz frequency shift","Two resonators cancel qubit coupling, then turn on with shift","From zero to 5 MHz: qubit coupling via double resonator","Qubit coupling switch: 50 MHz detuning flips from off to gate"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00086,"raw_usage":{"total_tokens":3540,"prompt_tokens":687,"completion_tokens":2853,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":431,"completion_tokens_details":{"reasoning_tokens":2757}},"tokens_in":431,"tokens_out":2853,"duration_ms":18872,"temperature":1.0,"reasoning_tokens":2757,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T00:04:05.369847+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the qubit-qubit coupling at the claimed switch-off point with a calibrated pulse sequence that corrects Z-pulse distortion and independently determine g12; if a nonzero anti-crossing gap or oscillation envelope persists there, or if the coupling at the 50 MHz-shifted point is below 5 MHz, the claimed cancellation would be called into question.","supporting_citations":[],"review_version":1}