{"id":"d3eb20ef-4afc-4a4d-9ca8-4b1d6f33687c","arxiv_id":"2509.03375","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A 'Late RWA' effective Hamiltonian that keeps counter-rotating drive terms reproduces measured ac Stark shifts and two-mode interactions where the conventional Early RWA approach fails.","lead":"This paper derives a better mathematical model for qubits and cavities driven by multiple microwave tones, fixing a known failure of earlier approximations. The new model matches experimental frequency shifts and interactions in circuit quantum electrodynamics, which could improve control of quantum information hardware.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The final RWA in the Late RWA derivation has no quantitative error bound and is never benchmarked against the full Hamiltonian; without such a check the claimed general validity of Heff is not established.","rationale":"The paper has real strengths: the ac Stark shift comparison in Fig. 1 is parameter-free and shows small stated deviations, and the sign change across resonance in Fig. 2 is a nontrivial feature that Early RWA cannot produce. These give independent support for a restricted claim. However, the headline claim is that Late RWA provides a general, correct effective Hamiltonian, and that requires the final RWA truncation to be controlled. The manuscript explicitly asserts the needed condition without quantifying it: the Supplement's only justification is that detunings are small enough to neglect superharmonic processes. No error bound, small-parameter estimate, or comparison against the untruncated dynamics is provided. This is the same weakest assumption the reader identified, so I agree with the reader. The most direct test is a numerical benchmark of Heff against the full Hamiltonian without the final RWA; this is feasible because the model is low-dimensional and the authors already perform QuTiP simulations. If that comparison succeeds in the tested regime, the concern is resolved; if not, the general claim is not supported. The fitted frequency offset nu_corr in the beam-splitting comparison and the lack of error bars are secondary because they would weaken only one of the two interaction demonstrations, whereas the missing no-RWA benchmark threatens the derivation itself. My concern therefore sharpens the existing CONDITIONAL verdict but does not move it.","tokens_in":13876,"tokens_out":12662,"duration_ms":120441,"concrete_test":"Use the authors' QuTiP setup to simulate the exact time-dependent Hamiltonian in the rotating frame (Eqs. S12-S14, before the final RWA) for the detuning sweep of Fig. 2, extracting the qubit Stark shift from the time evolution and comparing it against the spectrum of Heff. If the discrepancy near Δ≈-α exceeds the 1.3%/3.9% level claimed in Fig. 1, the final RWA is uncontrolled; if agreement persists, the concern is retired.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Heff in Eqs. (9)-(12) quantitatively reproduces driven qubit-cavity dynamics and is broadly applicable. For this claim, the final RWA after displacement must be accurate. The only validity condition stated in the paper appears in the Supplement after Eq. (S29): 'To apply the RWA, we assume drive detunings are small enough to neglect superharmonic processes, such as three-photon transitions (b†3) and similar terms.' No bound is derived, and the paper never compares Heff against the full Hamiltonian without the final RWA. The tested regime includes Fig. 2's detuning sweep to |Δ|≈300 MHz, which is not small compared with α≈230 MHz, and the avoided crossing near Δ≈-α is precisely where a truncated RWA can miss resonant level repulsion. If neglected superharmonic or slow-rotation terms are non-negligible there, the agreement of the Stark-shift curve could be coincidental rather than a valid approximation, and the claimed general applicability to arbitrary multi-tone driving would fail. The outlook section itself lists superharmonic processes as future work, confirming that this is an acknowledged gap rather than a closed result.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript derives an effective Hamiltonian for a qubit-cavity system under multi-tone off-resonant driving. The authors introduce a 'Late RWA' approach in which the rotating-wave approximation is applied only after a displacement transformation in the rotating frame, thereby retaining counter-rotating drive contributions. The resulting Hamiltonian (Eqs. (9)-(12)) contains diagonal Stark shifts, two-photon interactions (two-mode squeezing and beam-splitting), and counter-rotating corrections. The model is validated against experimental data from a previous paper by the same group, showing quantitative agreement for ac Stark shifts versus drive amplitude and detuning, two-mode squeezing chevrons, and beam-splitting dynamics. The paper claims that this model captures the near-resonant regime where the conventional 'Early RWA' approach fails.","tokens_in":14046,"tokens_out":5222,"duration_ms":47013,"significance":"If the central claims hold, the Late RWA effective Hamiltonian provides a practical and broadly applicable tool for simulating and designing driven bosonic interactions in circuit QED and related platforms. The key conceptual contribution is showing that counter-rotating drive terms can become resonant after displacement and that deferring the RWA captures near-resonant corrections missed by conventional displaced-frame treatments. The Stark-shift validation is a notable strength: for a single qubit drive, the model reproduces the measured shift to about 1.3% error without fitted parameters, and it captures the sign change across resonance and the avoided crossing at Δ ≈ -α. The explicit closed-form formulas make the model easily implementable, although no simulation code is shipped. The paper also usefully identifies the limits of the Early RWA method, which is widely used in the field.","major_comments":[{"comment":"The final RWA step is not quantitatively justified. The only validity condition given is that 'drive detunings are small enough to neglect superharmonic processes, such as three-photon transitions (b†3) and similar terms', but no error bound or numerical benchmark against the full Hamiltonian without this final RWA is provided. The tested regime includes detunings up to about 300 MHz, which is not small compared with the qubit anharmonicity α ≈ 230 MHz, and the avoided crossing near Δ ≈ -α is precisely where a truncated RWA can miss resonant level repulsion. Since the central claim is quantitative accuracy and general applicability to arbitrary multi-tone driving, the authors should either derive a bound on neglected terms or provide a comparison of Heff dynamics with an exact simulation of the full Hamiltonian (or the displaced Hamiltonian before the final RWA) over the parameter ranges used in Figs. 1-4. The outlook statement acknowledging superharmonic processes as future work confirms that this is an open issue rather than a closed result.","section":"Supplement, after Eq. (S29); main text Eq. (9)"},{"comment":"The derivation of the effective Hamiltonian is not shown in sufficient detail. The main text states that 'after performing the RWA, obtain' Eqs. (9)-(12), and the supplement says 'Expanding the quartic interaction ... we obtain' without displaying the intermediate algebra. Since Eqs. (9)-(12) are the main result, readers cannot verify the derivation or check which terms were kept and which were discarded. Please include the displaced expansion of the quartic term, the explicit identification of rotating vs. counter-rotating terms, and the selection rules used in the final RWA, or provide a symbolic-verification script (e.g., a Mathematica or Python notebook) as supplemental material.","section":"Supplement, around Eq. (S30)"},{"comment":"The beam-splitting simulation uses a fitted 'experimental correction' νcorr ≈ -5.02 MHz that shifts the cavity drive frequency. This is a free parameter not predicted by the model. While the use is disclosed, it means the beam-splitting data do not provide a parameter-free validation of Heff. Please quantify how sensitive the agreement in Fig. 4 is to νcorr and clarify which of the presented validations (Stark shifts, squeezing, beam-splitting) are fully parameter-free. The abstract's claim of 'quantitatively reproduces experimentally measured ac Stark shifts and captures key interactions' is supported by the Stark-shift data, but the beam-splitting demonstration would be more convincing with a sensitivity analysis or with νcorr predicted from the model.","section":"Main text, Fig. 4 and surrounding text"}],"minor_comments":[{"comment":"The shorthand notation is defined incorrectly: the text says 'ξ_i,2(t) = ∑_n ξ_i,1(t)' but the second sum should be over ξ_i,2(t). Please correct the typo.","section":"Supplement, after Eq. (S29)"},{"comment":"The horizontal axis label 'Detuning q [MHz]' should be 'Δ_q [MHz]' to match the text, and the y-axis label should be 'Stark shift [MHz]' with the units. The current label is ambiguous.","section":"Fig. 2 caption and axis labels"},{"comment":"The sentence 'Avoiding the conventional approximation a†e^{-iωt}+h.c., we retain the full cosine form' is confusing: the conventional approximation is typically to drop the counter-rotating term, not to use the specific complex-exponential notation. Please rephrase to clarify what the 'full cosine form' means and how Eq. (3) relates to it.","section":"Main text, Eq. (3) and surrounding text"},{"comment":"The symbol νcorr is introduced in the caption but is not defined in the main text. Please define it explicitly when the beam-splitting drives are first discussed.","section":"Main text, Fig. 4 caption"},{"comment":"The statement 'In [30], only the cavities are displaced, which does not capture the effects of off-resonance qubit driving' is too brief. Provide one or two sentences explaining the qualitative difference between displacing the cavity only and displacing both modes.","section":"Main text, paragraph after Eq. (11)"}],"recommendation":"major_revision","confidential_remarks":"The paper validates the model against data from the authors' earlier experiment [14], which is appropriate for this type of follow-up, but the beam-splitting comparison includes a fitted frequency offset. The main scientific risk is the unbenchmarked final RWA; if the authors can add a numerical benchmark against the full Hamiltonian and also make the derivation algebra available, the result would be solid. The topic fits the journal's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The Late RWA construction is a real contribution. The key idea—move to the rotating frame, displace, and only then apply the RWA—preserves counter-rotating drive terms that cancel the rotation of certain nonlinear terms, and that fixes the sign and magnitude of the ac Stark shift that the conventional Early RWA gets wrong. The data in Figs. 1–3 are convincing: the Stark shift curve vs detuning shows the sign change near zero and the avoided crossing near −α, and the two-mode squeezing and beam-splitting maps match the experiment well. This is a useful tool for circuit QED and other driven anharmonic systems.\n\nWhat the paper does well, beyond the central result, is that it provides a clear demonstration of where the standard approach fails and shows explicitly why the counter-rotating terms matter. The supplementary material gives the displacement amplitudes with dissipative corrections, which is a nice touch.\n\nThe soft spots are real but addressable. The derivation of Eqs. (9)–(12) is compressed: the main text says “performing the RWA” and the supplement says “Expanding … we obtain” without showing the algebra. A serious referee will want to see the expansion at least sketched. More importantly, the final RWA is never benchmarked against the full Hamiltonian. The paper states the assumption that drive detunings are small enough to avoid superharmonic processes, but gives no quantitative bound, and the detuning sweep goes to |Δ| ~ 300 MHz with α ~ 230 MHz—exactly the regime where the discarded terms may not be negligible. The stress-test point about the avoided crossing near Δ ≈ −α is well taken. A numerical comparison of Heff against the full (non-RWA'd) Hamiltonian for a few representative parameter points would settle this and should be in the paper. The beam-splitting validation uses a fitted offset ν_corr and the experimental data lack error bars; that lowers the evidential weight but does not undermine the central Stark shift result.\n\nOverall, the central claim is probably right, but the current manuscript is not complete as it stands. It deserves serious peer review—send it out, and let the referees push for the missing benchmark and derivation details.","headline":"Late RWA is a genuine fix for the ac Stark shift, but the missing benchmark against the full Hamiltonian keeps this from being a closed result.","tokens_in":14628,"tokens_out":3030,"would_cite":true,"duration_ms":27096,"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":"By deferring the rotating-wave approximation until after a displacement transformation, this paper derives an effective Hamiltonian that reproduces measured Stark shifts and driven interactions in a qubit-cavity system where the early-RWA…","keywords":["effective Hamiltonian","off-resonant driving","rotating-wave approximation","displacement transformation","circuit QED","ac Stark shift","two-mode squeezing","beam splitting"],"falsifier":"Tune a drive tone in the same qubit-cavity setup so that a discarded superharmonic process such as $b^{\\dagger 3}$ becomes resonant, then compare the measured transition rate with the prediction of Eqs. (9)--(12); the omitted term would make the effective Hamiltonian miss the resonance. A milder check is a Stark-shift-versus-detuning scan pushed well beyond $\\Delta_q\\approx -\\alpha$, where the \"small detunings\" assumption should eventually break down.","tokens_in":13640,"feed_emoji":"⚛️","tokens_out":8465,"duration_ms":74759,"temperature":0.7,"pith_summary":"Off-resonant multi-tone driving of a qubit-cavity system is hard to model: the standard recipe of moving to a displaced frame and then immediately applying the rotating-wave approximation predicts, for example, the wrong ac Stark shift of the qubit. This paper shows that the order of those two steps matters. Deferring the rotating-wave approximation until after the displacement keeps counter-rotating drive components that, together with anharmonic terms, produce slow, resonant contributions at small detunings, and the authors write down the resulting effective Hamiltonian explicitly. Tested against circuit-QED measurements, the model reproduces the measured Stark shifts to about one to four percent, the sign change of the shift across resonance, and the interference patterns of two-mode squeezing and beam-splitting gates.","feed_headline":"Deferring the rotating-wave approximation fixes qubit-cavity shifts","feed_subtitle":"By applying the RWA after displacement, the model matches measured Stark shifts and squeezing patterns.","key_machinery":"The machine that carries the argument is the time-dependent displacement transformation $U(t)=D_q[\\xi_q(t)]D_c[\\xi_c(t)]$, with displacement amplitudes split into co-rotating and counter-rotating parts, $\\xi_i(t)=\\xi_{i,1}(t)+\\xi_{i,2}(t)e^{2i\\omega_i t}$. The counter-rotating pieces are the ones the early-RWA approach discards; keeping them lets terms such as $b^{\\dagger 2}b$ rotate slowly at small drive detunings and contribute to the effective dynamics. The amplitudes are fixed by requiring the linear drive terms to cancel, with decay rates $\\kappa_q,\\kappa_c$ included so that the Lindblad dissipators remain invariant. Expanding the quartic Josephson term in the displaced frame and then applying the RWA produces Eqs. (9)--(12).","core_discovery":"The central claim is that the dynamics of a dispersively coupled qubit-cavity system under arbitrary off-resonant multi-tone drives is captured by the effective Hamiltonian of Eqs. (9)--(12). The derivation keeps the full cosine form of the drives, applies a displacement transformation whose amplitudes include both co-rotating ($\\xi_{i,1}$) and counter-rotating ($\\xi_{i,2}$) parts, and only at the final step applies the rotating-wave approximation. The resulting Hamiltonian contains a diagonal part with drive-induced frequency shifts $\\delta_q$ and $\\delta_c$, an interaction part $H_1$ describing displaced anharmonic and dispersive terms, and a counter-rotating correction $H_2$. The paper validates this construction by showing that its spectrum matches experimental ac Stark shifts, including the sign change near resonance and the avoided-crossing feature near $\\Delta_q\\approx -\\alpha$, and that its dynamics reproduce the measured populations in two-mode squeezing and beam-splitting protocols.","pith_inferences":["An implicit design rule, not stated in the paper, is to check any rapidly rotating term against every counter-rotating drive component before discarding it; applying that check systematically to sixth-order Josephson terms is the natural route to superharmonic transitions.","In platforms with weaker anharmonicity or stronger drives, the $H_2$ counter-rotating correction could grow relative to $H_1$, so the Late RWA ordering should matter in optomechanical and trapped-ion settings, not only in circuit QED.","A sharper validation than the ground-state Stark shift would be to measure the Fock-resolved dispersive shift under multi-tone drive; the model predicts explicit photon-number dependence through the $b^\\dagger b a^\\dagger a$ and $b^\\dagger a^\\dagger a$ terms, and a mismatch at higher $n$ would localize where the small-detuning assumption breaks."],"forward_implications":["The closed-form shifts $\\delta_q=-2\\alpha|\\xi_{q,1}|^2-\\chi|\\xi_{c,1}|^2$ and $\\delta_c=-2K_c|\\xi_{c,1}|^2-\\chi|\\xi_{q,1}|^2$ make cumulative Stark effects from complicated drive spectra predictable before running a master-equation simulation.","Choosing opposite detunings, $\\Delta_c=-\\Delta_q=\\Delta$, makes the two-photon term $b^\\dagger a^\\dagger$ resonant, enabling two-mode squeezing; choosing matched detunings, $\\Delta_c=\\Delta_q$, makes the exchange term $b a^\\dagger$ resonant, enabling beam-splitting.","Since the derivation only uses the quartic expansion of the Josephson cosine, the same effective-Hamiltonian construction transfers to any anharmonic oscillator, not just the transmon.","The model's agreement with the measured sign change across resonance means that Stark-shift-based calibration can use $H_{\\rm eff}$ to place operating points away from the avoided crossing near $\\Delta_q\\approx -\\alpha$.","The truncation that drops superharmonic terms such as $b^{\\dagger 3}$ is a stated boundary: transitions like $|0f\\rangle\\leftrightarrow|1g\\rangle$, which require such terms, are not covered by the present version."],"supporting_citations":[{"why":"supplies the experimental ac Stark shifts and the two-mode squeezing and beam-splitting data used to validate the effective Hamiltonian.","marker":"[14]"},{"why":"defines the conventional displaced-frame early-RWA approach whose predictions the paper shows to fail and must beat.","marker":"[15]"},{"why":"provides the normal-mode transformation that turns the capacitive coupling into the quartic Josephson expansion used in the derivation.","marker":"[49]"},{"why":"supports the open-systems step that keeps the Lindblad dissipators invariant under the displacement transformation.","marker":"[50]"},{"why":"shows a displacement treatment that displaces only the cavities, used as a contrast case that misses off-resonance qubit-drive effects.","marker":"[30]"}],"fun_headline_variants":["Delayed RWA matches qubit-cavity shifts and squeezing","Displacement-first theory reproduces driven qubit-cavity shifts","Effective Hamiltonian tames multi-tone drive errors in circuit QED","RWA after displacement captures Stark shifts and squeezing"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that every term discarded by the final rotating-wave step really rotates fast enough to average out, in particular superharmonic processes such as $b^{\\dagger 3}$; the paper assumes the drive detunings are small enough to suppress these but gives no quantitative boundary for that regime.","fun_headline_variants_meta":{"raw":{"variants":["Delayed RWA matches qubit-cavity shifts and squeezing","Displacement-first theory reproduces driven qubit-cavity shifts","Effective Hamiltonian tames multi-tone drive errors in circuit QED","RWA after displacement captures Stark shifts and squeezing"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000238,"raw_usage":{"total_tokens":1472,"prompt_tokens":868,"completion_tokens":604,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":533}},"tokens_in":484,"tokens_out":604,"duration_ms":5710,"temperature":1.0,"reasoning_tokens":533,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:30:13.422392+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Tune a drive tone in the same qubit-cavity setup so that a discarded superharmonic process such as $b^{\\dagger 3}$ becomes resonant, then compare the measured transition rate with the prediction of Eqs. (9)--(12); the omitted term would make the effective Hamiltonian miss the resonance. A milder check is a Stark-shift-versus-detuning scan pushed well beyond $\\Delta_q\\approx -\\alpha$, where the \"small detunings\" assumption should eventually break down.","supporting_citations":[{"cited_title":"Kudra, M","cited_arxiv_id":null,"evidence_quote":"supplies the experimental ac Stark shifts and the two-mode squeezing and beam-splitting data used to validate the effective Hamiltonian."},{"cited_title":"Campagne-Ibarcq, E","cited_arxiv_id":null,"evidence_quote":"defines the conventional displaced-frame early-RWA approach whose predictions the paper shows to fail and must beat."},{"cited_title":"Malekakhlagh, A","cited_arxiv_id":null,"evidence_quote":"provides the normal-mode transformation that turns the capacitive coupling into the quartic Josephson expansion used in the derivation."},{"cited_title":"Breuer and F","cited_arxiv_id":null,"evidence_quote":"supports the open-systems step that keeps the Lindblad dissipators invariant under the displacement transformation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"shows a displacement treatment that displaces only the cavities, used as a contrast case that misses off-resonance qubit-drive effects."}],"review_version":2}