{"id":"f31b9050-e312-45b7-a4d0-8e2e99a00288","arxiv_id":"2412.05886","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A noise-driven quantum-circuit refrigerator damps a microwave resonator's coherent state and reduces its effective temperature from roughly 300 mK to 130 mK with zero dc bias.","lead":"Applying broadband microwave noise far from the resonance of a superconducting resonator, the authors use a quantum-circuit refrigerator, a small tunnel junction, to damp coherent states and cool a thermal state from about 300 mK to 130 mK without any dc bias. The result suggests that waste heat or noise, rather than dedicated control lines, could power reset and refrigeration in future superconducting quantum processors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The qubit-spectroscopy thermometer is not validated at high QCR noise power, where line splitting, broadening, and background phase shifts are observed; the small ~0.7-photon cooling signal could be biased by line-shape distortion, so the 300-to-130 mK claim needs an independent check.","rationale":"The experiment contains real evidence for a plausible effect: raw spectral peak shifts are visible in Figs. 4 and 5, the noise-driven IV curves are modeled well in Fig. 2, and the underlying photon-assisted tunneling mechanism is well established. I do not think the paper should be rejected. However, the central quantitative claim, cooling from roughly 300 mK to 130 mK, is extracted by fitting qubit-spectroscopy peak heights under exactly the high-noise conditions where the paper itself reports line splitting, broadening, and background phase shifts. The reader's weakest assumption identifies this same unvalidated thermometer step, and it is the most load-bearing because a small systematic bias in the peak-height-to-population mapping could mimic the entire reported cooling. The secondary concern about Tqp = 60 mK versus the 248 to 250 mK used in the IV fits is real for the model comparison, but it would weaken the theory agreement rather than overturn the observed trend. A reanalysis of the raw spectra with a full line-shape model would settle whether the cooling is robust, so the conditional verdict remains appropriate.","tokens_in":12256,"tokens_out":11230,"duration_ms":131288,"concrete_test":"Reanalyze the raw qubit spectra of Figs. 4 and 5 with a multi-peak fit that includes power-dependent linewidths and a complex baseline phase, extracting the mean photon number from integrated peak areas rather than from peak heights with fixed line shapes; if the re-extracted mean photon number at the highest QCR noise power differs from the published value by more than about 0.15 photons (about 30 mK equivalent), the reported 300-to-130 mK cooling is not supported. This check can be performed on the existing data and directly tests whether readout distortion is the source of the apparent cooling.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing step is the conversion of qubit-spectroscopy peak heights into Fock-state populations at high QCR noise power. In Sec. III, Fig. 3(a), the authors report power-dependent line splitting, qubit linewidth broadening, a qubit-state-independent background phase shift, and loss of readout signal at strong noise drive. The photon-number extraction in Figs. 4(b) and 5(b) fits only the relative peak magnitudes of the qubit spectrum to Poisson/thermal distributions; no correction for these drive-induced line-shape distortions is described. If strong noise broadens peaks unevenly, tilts the baseline, or changes the calibration of peak height to population, the fitted mean photon number is biased even if the resonator population is unchanged. The reported thermal cooling is only from nbar approximately 0.92 (approximately 300 mK) to nbar approximately 0.22 (approximately 130 mK), a shift of about 0.7 photons, so a modest systematic error in the height-to-population mapping could account for the entire effect. The paper itself notes large fit uncertainty at higher photon numbers (Sec. V). This is the weakest link because it bears directly on the central quantitative claim; the free model parameters Tqp = 60 mK and nc = 0.92 affect only the theory comparison, not the observation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports an experimental study of a superconducting resonator coupled to a quantum-circuit refrigerator (QCR) driven by broadband noise rather than dc bias. The authors reconstruct resonator Fock-state populations from dispersive transmon spectroscopy and observe that resonant noise heats the resonator to a thermal distribution, while noise detuned far from the resonator and applied to the QCR dampens a coherent state and reduces the mean photon number of an initially thermal state from n̄≈0.92 (~300 mK) to n̄≈0.22 (~130 mK). They model the noise drive as a monochromatic ac bias at the noise center frequency and compare measured decay rates and steady-state populations with the model using Eq. (6) and Eq. (7).","tokens_in":12565,"tokens_out":6579,"duration_ms":65673,"significance":"If verified, the work demonstrates a qualitatively new operating mode for QCRs—refrigeration powered by noise only—and would support proposals for autonomous thermal machines and qubit reset driven by waste heat. The paper's strengths include direct spectroscopic evidence of noise-induced damping, quantitative IV-curve fits, and a transparent analytical model. However, the central quantitative claim (a ~0.7-photon reduction in thermal population) rests on qubit-spectroscopy peak-height fits at high noise power, where line-shape distortions are observed and not independently calibrated; the model comparison also uses parameters that appear chosen to match the data. These issues affect the strength of the cooling claim but not the existence of the damping effect.","major_comments":[{"comment":"The qubit-spectroscopy thermometer is not independently validated in the high-noise regime where the cooling effect is claimed. In Fig. 3(a) the authors report power-dependent line splitting, qubit linewidth broadening, a qubit-state-independent background phase shift, and loss of readout signal at strong noise drive. The photon-number extraction in Figs. 4(b) and 5(b) fits only the relative magnitudes of the spectral peaks to Poisson or thermal distributions, with no correction for these line-shape distortions and no systematic-error budget. Because the central cooling claim is a shift of only about 0.7 photons (from n̄≈0.92 to ≈0.22), a modest bias in the peak-height-to-population mapping due to uneven broadening or a tilted baseline could account for the entire effect. I recommend an independent calibration of the thermometer in the high-noise regime, for example by applying a known coherent drive to produce a reference Poisson distribution or by comparing with an alternative measurement of the resonator population such as the ac Stark shift or the resonator transmission linewidth.","section":"Sec. III, Figs. 3(a), 4(a), 5(a); Sec. V"},{"comment":"The model comparison in Fig. 5(b) uses nc=0.92 and Tqp=60 mK. The value nc=0.92 is essentially the zero-noise measured mean photon number, so the model is forced through the initial data point. Moreover, Tqp=60 mK differs from the quasiparticle temperatures Tqp=248 mK and 250 mK used to fit the IV curves in Figs. 2(b) and 2(d). If Tqp and nc are free parameters, the agreement in Fig. 5(b) is partly by construction; if they are independently determined, their derivation should be stated explicitly. The paper should also show how the predicted cooling curve shifts under the uncertainty in Tqp and nc, because this determines whether the data support a specific cooling magnitude or only a qualitative trend.","section":"Sec. V, Eq. (7), Fig. 5(b)"},{"comment":"The replacement of the finite-band noise drive by a monochromatic ac drive at the center frequency, with Vac=2√(2PNZ0), is a crude approximation that is not quantitatively validated. The IV-curve fits in Figs. 2(b) and 2(d) use the same model but require slightly different Tqp values (248 mK vs 250 mK), so the agreement does not constitute a stringent test of the monochromatic approximation. Since the predicted QCR-induced decay rate in Eq. (6) depends sensitively on the Bessel-function weights in Eq. (1), the authors should discuss the expected sensitivity to the finite noise bandwidth and, if possible, compare with a multi-tone or broadband calculation. Without this, the theoretical support for the extracted decay rates is weaker than implied.","section":"Sec. IV, Eq. (1) and Sec. V, Eq. (6)"}],"minor_comments":[{"comment":"The phrase \"AFM noiseform\" appears to be a typo; it should be \"AFM noise form\" or \"AFM noise waveform.\"","section":"Sec. III"},{"comment":"The sentence attributing the asymmetry in Fig. 3(d) to \"the asymmetric resonance peak of the resonator (data not shown)\" is unverifiable; either show the relevant resonance data or soften the attribution.","section":"Sec. III, Fig. 3(d)"},{"comment":"The bosonic population nQCR(V) is used in Eq. (7) but is not explicitly defined; after Eq. (8) it would be helpful to state nQCR = [exp(ℏωR/(kBTQCR))−1]^{-1}.","section":"Sec. V, Eq. (8)"},{"comment":"The color-bar axes for noise power are labeled only \"Noise power\" without units; the caption provides values in dBm, but the axis labels should include the units for clarity.","section":"Sec. V, Figs. 4(a) and 5(a)"},{"comment":"The paper notes that fit uncertainty is relatively large at higher photon numbers due to the monotonic shape of the thermal distribution, but this uncertainty is not quantified in the figure or text; adding representative error bars on the extracted points would strengthen the presentation.","section":"Sec. V, Fig. 5(b)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript fits the scope of the journal and reports a potentially significant effect. The main weakness is the lack of independent validation of the qubit-spectroscopy thermometer at high noise power, which is the basis for the quantitative cooling claim. The model-comparison parameters also need clarification. These issues are addressable with additional analysis or a targeted control experiment, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this paper shows that a QCR driven purely by broadband noise, far detuned from the resonator and with no dc bias, damps a coherent state and appears to cool a thermal state. The damping is directly visible in the qubit spectra — the heights of the Fock-state peaks shift with noise power in the expected direction. The cooling claim is less solid. The magnitude depends on converting those peak heights into photon numbers, and that conversion is not validated in the regime where the QCR is active.\n\nWhat's new: driving a single-junction QCR with noise instead of dc or coherent rf is a legitimate step beyond the prior art, including Refs. [21,22]. The paper's ac-drive model, borrowed from Ref. [37], captures the IV curves and the trend of the decay rate reasonably well. The authors are also honest about the rough spots — they mention line splitting, qubit linewidth broadening, a background phase shift, and larger fit uncertainty at higher photon numbers.\n\nThe soft spot is the thermometer. The qubit-spectroscopy readout is used to extract mean photon number by fitting relative peak heights to thermal or Poisson distributions. At high QCR noise power, the same spectra show power-dependent line splitting and broadening. The paper does not show that these distortions leave the height-to-population mapping unbiased. The reported cooling is only from about 0.9 photons to 0.2 photons — a shift of 0.7 photons. A modest systematic error in the peak-height calibration could account for all of it. The free parameters Tqp = 60 mK and nc = 0.92 are used in the model comparison but are not independently fixed, so the agreement in Figs. 4(b) and 5(b) is partly circular. None of this kills the qualitative observation — the raw damping is there — but the quantitative 130 mK claim should be treated as provisional until the readout is independently checked.\n\nWho is this for? People working on superconducting circuit reset, on-chip refrigeration, and autonomous quantum thermodynamics. It deserves a serious referee. The referee should ask for an independent calibration of the qubit-spectroscopy thermometer at high noise power, and for a sensitivity analysis showing how the extracted temperatures vary with Tqp and nc. If those checks hold up, this becomes a useful reference; if not, the qualitative damping still stands but the headline cooling number needs to be walked back.\n\nRecommendation: send it to peer review, with the thermometer validation as the key revision.","headline":"Noise-driven QCR damping is a real, visible effect; the 300-to-130 mK number rests on a thermometer that isn't validated at high noise power.","tokens_in":13067,"tokens_out":1834,"would_cite":true,"duration_ms":19920,"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":"Broadband noise alone can power a quantum-circuit refrigerator, damping a coherent state and cooling a superconducting resonator from about 300 mK to 130 mK.","keywords":["quantum-circuit refrigerator","photon-assisted tunneling","superconducting resonator","noise-driven cooling","transmon qubit readout","Fock-state populations","normal-metal-insulator-superconductor junction","autonomous thermal machines"],"falsifier":"Measure the resonator's mean photon number with an independent readout that does not rely on qubit-spectroscopy peak heights—for example, the resonator's own linear transmission—while sweeping the same far-detuned noise power; if the independently extracted temperature does not drop from roughly 300 mK to 130 mK, the reported cooling is a readout artifact.","tokens_in":12055,"feed_emoji":"❄️","tokens_out":9400,"duration_ms":81868,"temperature":0.7,"pith_summary":"This paper reports an experimental demonstration that a quantum-circuit refrigerator (QCR)—a normal-metal–insulator–superconductor tunnel junction galvanically attached to a superconducting microwave resonator—can be driven purely by broadband noise, with no dc voltage across the junction. When the noise band is tuned far from the resonator frequency, the QCR absorbs photons from the resonator and damps a coherent state, and it reduces the effective temperature of a thermal state from roughly 300 mK to 130 mK. The authors show that the same noise applied on resonance heats the resonator into a thermal distribution, so the off-resonant drive acts like a filtered thermal bath rather than a simple heater. If the result holds, it opens a route to resetting superconducting qubits and building autonomous quantum heat engines powered by environmental thermal noise.","feed_headline":"Noise alone cools a quantum circuit from 300 mK to 130 mK","feed_subtitle":"A far-detuned noise drive damps a microwave resonator with no dc bias, pointing to waste-heat-powered qubit reset.","key_machinery":"The load-bearing element is the quantum-circuit refrigerator (QCR): a normal-metal–insulator–superconductor tunnel junction whose inelastic photon-assisted quasiparticle tunneling removes photons from the galvanically coupled resonator. The theoretical machinery is the ac-drive model of Ref. [37], in which the finite-bandwidth noise is approximated by a sinusoidal voltage $V_{\\rm ac}\\cos(\\omega_{\\rm ac}t)$ at the noise center frequency; the resulting Fock-state transition rates (Eq. (1)) feed the QCR-induced decay rate $\\gamma_{\\rm QCR}(V)$ in Eq. (6) and the weighted-average steady-state population $\\bar{n}(V)=[n_{\\rm QCR}(V)\\gamma_{\\rm QCR}(V)+n_c\\gamma_c]/[\\gamma_{\\rm QCR}(V)+\\gamma_c]$ in Eq. (7). This lets the authors predict the observed noise-power dependence from device parameters such as tunneling resistance, gap, Dynes parameter, and quasiparticle temperature, rather than treating the cooling as a free fit.","core_discovery":"The paper's central claim is that a quantum-circuit refrigerator attached to a superconducting microwave resonator can be operated with zero dc bias and driven purely by broadband noise, provided the noise band is far detuned from the resonator frequency. Using a transmon qubit to resolve the resonator's Fock states, the authors observe that increasing the off-resonant noise power damps a coherent state and reduces the effective temperature of a thermal state from roughly 300 mK to 130 mK. The same noise applied on resonance instead heats the resonator into a thermal distribution, confirming that the off-resonant drive acts as a filtered thermal bath rather than as a direct heater. The authors model the noise as an equivalent sinusoidal drive at the noise center frequency and find that the measured decay rates and steady-state populations agree with the photon-assisted quasiparticle-tunneling rates of their QCR model.","pith_inferences":["Extending beyond the paper, the required noise power could be lowered by raising the resonator impedance and mode frequency, eventually allowing an on-chip resistor thermalized at a higher cryostat stage to serve as the noise source.","The same zero-dc-bias dissipation mechanism could be applied directly to qubit initialization, replacing coherent reset pulses with a thermal-noise-driven reset element.","Because the steady-state population in Eq. (7) is a weighted average of two baths, the noise-driven QCR offers a way to tune a resonator's effective bath temperature continuously, which could be used to study quantum heat engines and thermal machines."],"forward_implications":["A qubit-reset element for a superconducting processor can in principle be powered by noise alone, removing the dc bias line and its associated control electronics.","The noise source acts as a tunable thermal bath: resonant noise heats the resonator, while far-detuned noise cools it, giving a single device both heating and cooling functionality.","Because the measured decay rates match the photon-assisted-tunneling model, device parameters such as junction resistance and gap can be used to engineer the cooling rate without empirical tuning.","The demonstrated reduction from roughly 300 mK to 130 mK places the resonator closer to the quantum ground state, which is the regime needed for high-fidelity qubit operations."],"supporting_citations":[{"why":"Supplies the photon-assisted quasiparticle-tunneling rate equations and the elastic-transition approximation that underpin Eqs. (1)-(3).","marker":"[26]"},{"why":"Provides the ac-drive model in which the finite-band noise is replaced by a sinusoidal drive at the noise center frequency.","marker":"[37]"},{"why":"Gives the sample design, the QCR-resonator measurement protocol, and the baseline decay rates to which the noise-driven QCR is compared.","marker":"[31]"},{"why":"Establishes the strong-dispersive qubit-spectroscopy readout that resolves Fock-state populations and provides the photon-number extraction used in Figs. 3-5.","marker":"[45]"},{"why":"Introduces the quantum-circuit refrigerator device and its operation principle, which this work drives with noise instead of dc or coherent drive.","marker":"[25]"},{"why":"Documents earlier QCR-induced decay rates in a resonator and supplies model parameters and expected magnitudes used in the comparison.","marker":"[30]"}],"fun_headline_variants":["Off-resonant noise cools a quantum circuit to 130 mK","Noise-only cooling halves resonator temperature, no dc bias","Quantum fridge powered by off-resonant noise, not electricity","Noise drive chills microwave resonator from 300 to 130 mK","Harnessing waste heat: noise cools qubit reset circuit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reported cooling rests on the assumption that the heights of the qubit-spectroscopy peaks still faithfully encode the resonator's photon-number distribution when the noise drive is strong enough to distort the spectrum.","fun_headline_variants_meta":{"raw":{"variants":["Off-resonant noise cools a quantum circuit to 130 mK","Noise-only cooling halves resonator temperature, no dc bias","Quantum fridge powered by off-resonant noise, not electricity","Noise drive chills microwave resonator from 300 to 130 mK","Harnessing waste heat: noise cools qubit reset circuit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000265,"raw_usage":{"total_tokens":1577,"prompt_tokens":884,"completion_tokens":693,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":500,"completion_tokens_details":{"reasoning_tokens":603}},"tokens_in":500,"tokens_out":693,"duration_ms":6714,"temperature":1.0,"reasoning_tokens":603,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T20:14:31.390827+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the resonator's mean photon number with an independent readout that does not rely on qubit-spectroscopy peak heights—for example, the resonator's own linear transmission—while sweeping the same far-detuned noise power; if the independently extracted temperature does not drop from roughly 300 mK to 130 mK, the reported cooling is a readout artifact.","supporting_citations":[{"cited_title":"Silveri, H","cited_arxiv_id":null,"evidence_quote":"Supplies the photon-assisted quasiparticle-tunneling rate equations and the elastic-transition approximation that underpin Eqs. (1)-(3)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the ac-drive model in which the finite-band noise is replaced by a sinusoidal drive at the noise center frequency."},{"cited_title":"Viitanen, T","cited_arxiv_id":null,"evidence_quote":"Gives the sample design, the QCR-resonator measurement protocol, and the baseline decay rates to which the noise-driven QCR is compared."},{"cited_title":"Viitanen, M","cited_arxiv_id":null,"evidence_quote":"Documents earlier QCR-induced decay rates in a resonator and supplies model parameters and expected magnitudes used in the comparison."}],"review_version":1}