{"id":"2bac6e32-af90-4481-a450-d575c7c8ac8d","arxiv_id":"2608.00162","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A driven transmon-cavity system pumps photons into a microwave cavity at a topologically quantized rate, reaching ~7 photons with sub-Poissonian statistics.","lead":"A superconducting qubit in a microwave cavity was driven so that exactly one photon enters the cavity every cycle, taking the cavity from empty to about seven photons. The paper reports the first direct observation of this topologically quantized photon pump and shows the produced states are non-classical for the first few photons.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No independent χ1 calibration: the P/Pq≈1 headline scales directly with the numerically computed single-photon shift χ1=−1.74 MHz and is not covered by the fit error bars.","rationale":"The reader's CONDITIONAL verdict is warranted. The paper's internal consistency, phase-diagram agreement, and Fock-state evidence are real and support the phenomenon. But the central 'quantized rate' statement rests on an absolute photon-number calibration that is numerical rather than independently measured. This is an addressable, non-fatal issue: it does not contradict the internal logic, and the proposed calibration could resolve it. The reader's weakest assumption identifies the same load-bearing point, so no verdict change is needed.","tokens_in":23080,"tokens_out":13907,"duration_ms":176038,"concrete_test":"At the standard operating point (ωq0=4.90 GHz, Δqc=−140 MHz), prepare the storage cavity in a single-photon Fock state using a calibrated qubit-cavity swap (or by one pump cycle and independent verification via the resolved spectrum at −40 MHz), and measure the resulting qubit AC Stark shift. Compare this measured χ1 with the QuTiP value −1.74 MHz, then refit Fig. 3's P/Pq using the measured χ1. If P/Pq in the topological regime shifts by more than the reported error bars, the quantized-rate claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central result of Fig. 3—P/Pq≈1 throughout the topological regime—is not a direct topological measurement but a fit to n(mT)=(P/κ)(1−e^{−κmT}), where n(mT) is obtained from the qubit AC Stark shift divided by χ1. As stated in S5D, χ1=−1.74 MHz is 'determined from numerical calculation with QuTip and the chip parameters in Table S1'; no independent experimental calibration of χ1 at the −140 MHz operating point is presented. Because the fit uses the first five cycles, where κmT≈0.08, the extracted P is essentially the early-time slope of n(mT); a multiplicative error in χ1 maps directly into P/Pq. The reported error bars are only fit uncertainties and do not include this calibration error, nor the uncertainty in the chip parameters (g, α, Δqc) entering QuTiP. Since the headline claim is a quantized rate P/Pq=1, a systematic error of even 10–20% in χ1 would move P/Pq away from unity by more than the visual precision of Fig. 3. This is the load-bearing step: without an independently measured χ1, the absolute quantization cannot be distinguished from a calibrated-but-wrong scale.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a circuit-QED realization of a topological photon pump. A transmon qubit coupled to a microwave cavity is driven by a semicircular effective field B(t) synthesized from flux and microwave control. Starting from vacuum, the authors observe a linear-to-saturating increase of the qubit AC Stark shift with number of drive cycles, interpret this as n(mT) photons in the cavity, map a phase diagram in (B, Δ) whose resonance dips match an analytic formula, extract a pump rate P from a loss-model fit, and find P/Pq ≈ 1 in the topological regime. They also present photon-number-resolved spectroscopy showing sub-Poissonian states after the first few cycles, and a bidirectional drive experiment. Supplemental material derives the semiclassical Floquet quasi-energy spectrum, the phase boundaries, the Chern-number picture, and the maximum photon number.","tokens_in":23342,"tokens_out":8670,"duration_ms":92613,"significance":"If the calibration concern is resolved, these data constitute the first direct observation of a topological photon pump in the quantum regime and a meaningful step toward robust Fock-state preparation. The analytic phase diagram (Eq. S27), the Floquet/Chern classification in Sec. S2, and the extensive parameter sweeps are substantial contributions. The experiment is an external test of the authors' earlier theoretical proposal (Ref. 26), which counts as independent support rather than circularity. The main caveat is that the absolute photon-number scale, and hence the quantized-rate claim, rests on a numerically computed χ1 without independent experimental calibration.","major_comments":[{"comment":"The absolute scale of n(mT) is set by χ1 = −1.74 MHz, obtained from a QuTiP calculation with the Table S1 parameters, with no independent experimental calibration at the −140 MHz operating point. Since P is extracted by fitting Eq. S32 to the first five cycles, a multiplicative error in χ1 enters P/Pq almost directly. The error bars in Fig. 3 are only fit uncertainties. Please add an independent χ1 measurement (e.g., a displacement-induced shift) or propagate the uncertainty in g, α, Δ_qc and show that the resulting P/Pq interval contains unity. Without this, the headline 'P ≈ Pq' cannot be distinguished from a miscalibrated scale.","section":"Sec. S5D / Fig. 3"},{"comment":"The claim that the produced states are 'demonstrably non-classical' is supported visually by comparison with Fock and coherent distributions, but no quantitative nonclassicality witness is given. Reporting a Mandel Q parameter, g^(2)(0), Wigner negativity, or a Fock-state fidelity with uncertainty would make the claim precise. If a witness is not available, the wording should be weakened.","section":"Fig. 4 / Abstract"}],"minor_comments":[{"comment":"References [25] and [30] are the same paper (Nathan, Martin, and Refael, Phys. Rev. B 99, 094311 (2019)). Consolidate to avoid duplicate citation.","section":"References [25] and [30]"},{"comment":"The sentence 'We observe P≈Pq everywhere else in the topological regime' is stronger than the data shown in Fig. 3, which covers three values of Δ and a limited window of B. Qualify the claim to 'for all measured configurations' or show the full dataset.","section":"Main text, 'P≈Pq everywhere else'"},{"comment":"The upper and lower horizontal axes in Fig. 4 are not clearly distinguished in the figure as printed. A legend explaining the relationship between 'Frequency shift (χ1)' and 'exact qubit frequency shifts' would improve readability.","section":"Fig. 4 axes"},{"comment":"The sentence 'χ1 is smaller than the linewidth of the qubit' is ambiguous; the relevant linewidth is the measured spectroscopy linewidth, not the qubit decay rate. Please specify.","section":"Sec. S5D"}],"recommendation":"major_revision","confidential_remarks":"I am sympathetic to the result: the device, the protocol, and the theoretical framework are well matched, and the qualitative signatures (direction reversal, phase-diagram dips, Fock-state response) are convincing. The main obstacle is the absolute photon-number calibration. If the authors can supply an independent χ1 measurement or a fully propagated systematic uncertainty analysis, the quantitative claim of quantization would be established. Without that, the central P/Pq ≈ 1 result remains vulnerable to a shift of the photon-number scale."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, you should know two things. First, this is a real experimental milestone: a topological photon pump in circuit QED that starts from vacuum, adds roughly one photon per cycle, and shows sub-Poissonian statistics and bidirectional current. Second, the headline 'quantized pump rate' is not measured directly; it comes from a fit to the first five cycles, and the absolute photon scale is set by a numerically computed single-photon shift χ1. That is the load-bearing seam.\n\nWhat is genuinely new: prior topological pumps were spatially extended or proof-of-principle; this is the first direct observation in a cavity starting from vacuum. The paper also maps out a phase diagram with analytical boundaries and resonances, and the agreement between the measured nfinal and simulation in Fig 2(c) is nontrivial — the simulation uses independent chip parameters, so that match is an indirect check on the χ1 scale. The bidirectional pumping data in S1 is a nice topological signature, and the Fock-state-like spectra in Fig 4 support the non-classicality claim.\n\nWhere it is soft: The extraction of P in Fig 3 indeed has error bars that only reflect the fit uncertainty, not the systematic uncertainty in χ1 (or in g, α, Δqc entering QuTiP). The stress-test is right that a 10–20% error in χ1 would move P/Pq off unity by more than the visual precision. However, the paper actually contains two pieces of indirect calibration that soften this: the n(mT)×T collapse in Fig S8(c) against the P=Pq line, and the absolute match of nfinal to simulation. Neither is a clean independent measurement, but both suggest χ1 is close to correct. Still, the authors should either provide a direct χ1 calibration (e.g., from a resolved Fock-state spectrum at closer detuning on the same device) or release the raw n(mT) traces and analysis scripts so referees can propagate the uncertainty. The automatic rejection criteria are described, but their effect on the fitted P is not quantified; a simple robustness check (varying the 2χ1 threshold) would address this.\n\nBottom line: This is a serious paper with a mature theory and credible data. The central claim is defensible, but the word 'quantized' currently carries more weight than the absolute calibration supports. A referee should be told to focus on S5D and ask for the calibration or raw data. I would send it to peer review; it deserves careful attention.","headline":"First convincing quantum-regime topological photon pump, but the P/Pq≈1 headline rests on a numerically calibrated χ1 and fit error bars; still deserves refereeing.","tokens_in":23846,"tokens_out":4251,"would_cite":true,"duration_ms":51329,"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 the first direct observation of a topological photon pump: a transmon qubit coupled to a microwave cavity adds one photon to the cavity per drive cycle, at a rate fixed by the cycle period and insensitive to control-field","keywords":["topological photon pump","spectral flow","Thouless pump","cavity QED","superconducting qubit","Fock state","quantized pumping","circuit QED"],"falsifier":"Measure the cavity photon number after m cycles with a technique independent of the qubit's dispersive shift—for example, direct heterodyne detection of the cavity field or Wigner tomography—at the same Δ=101 MHz, B=120 MHz, ω=5 MHz settings; if the inferred photon numbers and P/Pq differ from 1 beyond the error bars, the quantized-rate claim fails.","tokens_in":22956,"feed_emoji":"⚛️","tokens_out":5565,"duration_ms":57409,"temperature":0.7,"pith_summary":"Topological pumps should move charge or energy at rates fixed by a band invariant, unaffected by the shape of the drive. This paper reports the first direct observation of such a pump acting on photons: a transmon qubit coupled to a microwave cavity is driven through a semicircular cycle, and each period adiabatically moves the system from |n,g⟩ to |n+1,g⟩, so one photon enters the cavity per cycle. Starting from vacuum, the authors observe up to ≈7 photons, an extracted pump rate P matching the quantized value Pq=1/T across the topological regime for different field amplitudes and detunings, and sub-Poissonian (Fock-like) cavity statistics for the first few cycles. The result matters because it offers a robust, calibration-insensitive way to prepare non-classical photon states and to reset cavities, in contrast to fine-tuned control protocols.","feed_headline":"One photon per cycle: first topological photon pump observed","feed_subtitle":"Driven qubit–cavity device pumps vacuum to ~7 photons at a quantized rate; early states are non-classical.","key_machinery":"The central object is the spectral flow of the instantaneous eigenstates of the driven Jaynes-Cummings Hamiltonian. During the vertical portion of the drive, polariton number is conserved, protecting exact level crossings between pumping (|n,g⟩→|n+1,g⟩) and depumping (|n,e⟩→|n−1,e⟩) branches; during the semicircular portion, the qubit changes state while the photon number stays fixed. This one-rung-up-per-cycle winding is equivalent to a nonzero Chern number of the dressed Floquet states on the (θ1,θ2) torus, and it fixes the pump rate regardless of the details of B(t) so long as the vertical stroke is ungapped. An analytic phase diagram gives the boundary B_min(Δ)=(Δ²−g²)/Δ, resonance detun","core_discovery":"The central claim is that in a circuit-QED system described by H_pump(t)=Δ a†a + g(a†σ−+aσ+)+ (1/2) B(t)·σ, with B(t) tracing a semicircle in the (Bx,Bz) plane, the instantaneous spectrum of the coupled qubit-cavity system winds by exactly one photon rung per drive period in the regime κ≪ω<g≪Δ,B. Because of the spectral flow, the state |n,g⟩ adiabatically connects to |n+1,g⟩ over each cycle, so starting from vacuum the cavity gains one photon per cycle with quantized rate Pq=1/T. The paper reports the first experimental observation of this: photon numbers growing from vacuum to ≈7, an extracted P/Pq ≈ 1 independent of Δ and B in the topological regime, near-zero pumping in the trivial regime","pith_inferences":["A direct test of the protection mechanism: deliberately distort the semicircular segment by adding a random or offset shape while keeping the vertical stroke intact and check P/Pq remains 1; the paper's symmetry argument predicts stability, and its offset simulations predict degradation when the vertical stroke itself is offset.","Because the quantized rate is a property of the spectral flow rather than of the specific transmon nonlinearity, the same protocol should transfer to other cavity-QED platforms (optical cavities coupled to atoms or Rydberg ensembles) whenever the effective field geometry can be engineered; the paper's outlook gestures this way but does not demonstrate it.","The superposition statement in the outlook—qubit in |g⟩+|e⟩ producing superpositions of Fock states with very different photon numbers—suggests the pump could be used to synthesize approximate cat states; a concrete next experiment would be to initialize the qubit in an equal superposition and perform Wigner tomography after a few cycles.","The paper's P/Pq fit assumes a simple exponential saturation n(mT)=(P/κ)(1−e^{−κmT}); at low ω the discrete photon-arrival timing causes deviations from this form. A more refined model that accounts for stroboscopic injection could yield an even cleaner quantization test and would be worth checking against the ω=1–10 MHz data."],"forward_implications":["If the central claim is right, a single drive period transfers exactly one photon into the cavity from vacuum, so the pump can serve as a calibration-free photon source: the rate is set by the clock period T, not by pulse shapes.","The same robustness means the pump can prepare approximate Fock states from vacuum; the observed sub-Poissonian statistics to ⟨n⟩≈3 are limited by cavity loss, and millisecond-lifetime cavities should reach much larger Fock states.","Reversing the drive direction reverses the photon current, depumping the cavity; this gives an on-demand reset to vacuum that is independent of the initial cavity state.","The pump remains quantized under deformations of the semicircular segment of the protocol, while transverse fields during the vertical stroke destroy pumping—identifying exactly which control imperfections matter.","The pump fails in the small-detuning regime due to qubit backaction on the cavity photon number, so the large-Δ regime, not the previously proposed small-Δ regime, is the practical one for preparing non-classical states from vacuum."],"fun_headline_variants":["First topological photon pump observed in circuit QED","Quantum pump: one photon added per cycle, experiment shows","Topological photon pump demonstrated with qubit-cavity","Quantized photon pump realized, first time in experiment"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The absolute photon numbers and extracted pump rate assume the numerically computed single-photon AC Stark shift χ1 = −1.74 MHz and the measured cavity decay rate κ are correct; if either calibration is off, all quoted n and P/Pq values scale and the P≈Pq conclusion is weakened.","fun_headline_variants_meta":{"raw":{"variants":["First topological photon pump observed in circuit QED","Quantum pump: one photon added per cycle, experiment shows","Topological photon pump demonstrated with qubit-cavity","Quantized photon pump realized, first time in experiment"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000118,"raw_usage":{"total_tokens":884,"prompt_tokens":673,"completion_tokens":211,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":417,"completion_tokens_details":{"reasoning_tokens":147}},"tokens_in":417,"tokens_out":211,"duration_ms":3202,"temperature":1.0,"reasoning_tokens":147,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T01:07:07.223705+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the cavity photon number after m cycles with a technique independent of the qubit's dispersive shift—for example, direct heterodyne detection of the cavity field or Wigner tomography—at the same Δ=101 MHz, B=120 MHz, ω=5 MHz settings; if the inferred photon numbers and P/Pq differ from 1 beyond the error bars, the quantized-rate claim fails.","supporting_citations":[],"review_version":1}