{"id":"0c03750b-9406-45b6-a7c3-eb6b5ff99b0c","arxiv_id":"2607.09178","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"Operating an MKID near bifurcation with the tone below resonance yields ~10× optical responsivity via readout-current positive feedback, at the cost of reduced bandwidth.","lead":"MKID detectors can be made about ten times more responsive to light by parking the readout tone just below resonance near the bifurcation point, using the detector’s own current nonlinearity as positive feedback. The same knob also shrinks bandwidth, so the paper maps the full trade-off for instrument designers.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified beyond the reader's already-flagged reactive-only modeling choice.","rationale":"The paper's strongest claim is an experimentally realized, theoretically mapped positive-feedback operating regime for MKIDs. The analytics (critical-point coordinates, feedback factor 1−∂Ẽ/∂x, eigenvalues s±) are standard Kerr-cavity results applied cleanly to detector coordinates (x0,Pg) and (x,xr). The laboratory evidence—power-sweep hysteresis matching the model, multi-resonator maps showing >10× responsivity near the critical point, and noise spectra in which both intrinsic white level and an extrinsic LED tone rise while bandwidth falls—directly corroborates the predicted trade-off. The only modeling idealization that could quantitatively alter the gain is the reactive-only assumption already identified by the reader; the paper itself notes that dissipative components should be included in future work and that no dissipative optical response was resolved above uncertainty. Because that caveat is already on the table and does not overturn the measured enhancement under the conditions tested, no further load-bearing concern is warranted. Verdict remains ACCEPT.","tokens_in":14223,"tokens_out":544,"duration_ms":6053,"concrete_test":"Re-fit the high-power spectrum of Fig. 8 with a two-pole model that includes both τ_r (from Eq. 20) and τ_qp; if the extracted low-frequency gain relative to the low-power spectrum remains ≥10 and the LED-to-noise ratio is preserved, the reactive-feedback claim is confirmed even with residual dissipative effects present.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (positive-feedback enhancement of optical responsivity by ~10\times near bifurcation, with corresponding bandwidth reduction) is supported by the Kerr-cavity steady-state and linearized dynamics (Eqs. 6, 14, 18, 20), the control-space and driven-state maps (Figs. 6–7), and the laboratory frequency-noise spectra that show the white level rising together with an extrinsic LED tone while the roll-off falls (Fig. 8). The reader's weakest assumption—neglect of dissipative nonlinearity and the fixed β=20—is already the softest modeling choice the paper itself flags (Sec. II.B, II.D). No additional load-bearing inconsistency or unsupported leap is required for the claim to hold under the stated reactive approximation. The measured spectra and multi-resonator average in Fig. 7 already provide independent experimental corroboration that the reactive feedback picture is adequate for these devices.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript treats readout-current nonlinearity in MKIDs as a controllable operating resource rather than a power limit. Using a Kerr-cavity input–output description, it shows that placing the generator tone below the driven resonance produces positive feedback between stored energy and resonant-frequency shift. Near the bifurcation (critical) point this feedback enhances optical responsivity by a measured factor of ~10 relative to the undriven value while lengthening the driven resonator relaxation time (reducing bandwidth). The authors derive the steady-state photon number, critical-point coordinates, small-signal feedback factor, and linearized eigenvalues; map the resulting responsivity and time-constant surfaces in both control-space (x0, Pg) and driven-state (x, xr) coordinates; and corroborate the predictions with fixed-frequency power sweeps, multi-resonator averages, and frequency-noise spectra that include an extrinsic LED tone.","tokens_in":14528,"tokens_out":1027,"duration_ms":17638,"significance":"If the result holds under the stated reactive approximation, the work supplies a practical, easily selected bias point that can improve MKID sensitivity whenever non-intrinsic noise (amplifier, TLS, etc.) dominates, at the cost of bandwidth. The operating-state maps and the explicit link between the same feedback factor and both gain and critical slowing are concrete, reusable contributions for array design and multiplexing. Strengths include a derivation that follows standard Kerr-cavity theory, independent experimental checks (noise floor and LED signal rise together while roll-off falls), and consistency across multiple resonators and materials. The paper itself flags the principal modeling limitation (reactive-only treatment), which keeps the claim falsifiable.","major_comments":[{"comment":"Sec. II.B and the linearized dynamics (Eq. 20): the analysis explicitly neglects nonlinear losses and dissipative quasiparticle generation by the readout current. While the measured spectra and multi-resonator average in Fig. 7 support reactive dominance for the devices studied, near the critical point even a modest dissipative component can shift the loop-gain zero, alter the eigenvalues, and change the attainable gain. A quantitative upper bound extracted from the same power-sweep or noise data (or a short appendix including two-photon loss) would make the claimed factor-of-10 enhancement more robust.","section":"Sec. II.B, Eq. 20"},{"comment":"Sec. II.D, Eq. 18–19: the optical-responsivity ratio is evaluated with a fixed reactive-to-dissipative ratio β = 20 chosen because no dissipative optical response was resolved. The text notes that smaller β would further enhance responsivity at x = 0, yet the maps and the 10\times claim are presented for this single value. A brief sensitivity plot versus β (or an experimental bound) would clarify how much of the reported enhancement is model-dependent.","section":"Sec. II.D, Eqs. 18–19"}],"minor_comments":[{"comment":"Fig. 7 caption and surrounding text introduce the conventional nonlinearity parameter a and its critical value a_c = 4√3/9 without an explicit definition in the main text; a short sentence linking a to Ẽ or n_ph would improve readability for non-specialists.","section":"Fig. 7"},{"comment":"Eq. 6 / A2 and the three-port schematic (Fig. 2) are clear, but the mapping from the side-coupled S21 geometry to the one-port Kerr cavity is only sketched; a one-line statement of the port identification would remove any residual ambiguity.","section":"Sec. II.B, Fig. 2"},{"comment":"In Fig. 8 the single-pole Lorentzian fit is noted to deviate at high power; stating the extracted τ_r / τ_qp ratio (or showing a two-pole residual) would make the critical-slowing interpretation more quantitative.","section":"Fig. 8"},{"comment":"Minor typographical inconsistencies appear in the quality-factor subscripts (Q_r vs. Qr) and in the units of the LED peak (37 Hz); a uniform pass would polish the manuscript.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The central claim is experimentally well-supported under the reactive approximation the authors themselves emphasize. The two major comments are constructive rather than blocking; once addressed (even briefly) the paper is ready for acceptance. Scope and novelty fit a condensed-matter / detector-physics journal well."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The useful takeaway is simple: put the tone below the driven resonance and park near bifurcation, and the current-dependent kinetic inductance gives you positive feedback that multiplies optical responsivity by ~10 while stretching the resonator time constant. They map the whole (x0, Pg) and (x, xr) space so you can dial the trade-off.\n\nWhat is actually new is the control-space framing and the measured maps, not the underlying physics. Nonlinear kinetic inductance, bifurcation, and Kerr-cavity input–output are already in Swenson, de Visser, Anferov, Yurke & Buks, etc. The paper’s contribution is treating the readout current as a deliberate feedback bias rather than a power ceiling, deriving the small-signal optical responsivity ratio (their Eq. 18) and the linearized eigenvalues (Eq. 20), then showing the corresponding rise in white noise level and drop in roll-off on real spectra (Fig. 8), including an extrinsic LED tone that scales with the floor. The multi-resonator average in Fig. 7 and the match of power-sweep hysteresis to the steady-state photon-number equation look solid. Citations are appropriate; the self-cites are apparatus context, not circular.\n\nSoft spots are the ones they already flag. They keep only the reactive nonlinearity and set β = 20 because they could not resolve a dissipative optical response. If dissipative quasiparticle generation by the readout current becomes important near the cusp, the feedback factor and the claimed gain will move. That is a modeling limitation, not a contradiction with their data on these TiN devices. Bandwidth is traded for low-frequency gain, which they state clearly; the mode is useful when amplifier or TLS noise dominates, not when you need the full quasiparticle bandwidth.\n\nThis is for people who actually bias MKID arrays and care about multiplexing density or non-intrinsic noise floors. The math is standard Kerr cavity, the lab spectra corroborate the prediction, and the maps are immediately usable. I would send it to peer review without hesitation; a referee can push on the dissipative terms and on how close to the cusp one can sit in a real multiplexed system. Worth reading if you work these detectors.","headline":"Solid subfield paper: known Kerr nonlinearity reframed as a selectable positive-feedback bias, with a clean ~10\times optical-responsivity measurement and usable operating maps.","tokens_in":15122,"tokens_out":541,"would_cite":true,"duration_ms":7037,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Placing the MKID readout tone just below the driven resonance near bifurcation multiplies optical responsivity by about ten while stretching the resonator time constant.","keywords":["microwave kinetic inductance detectors","kinetic inductance nonlinearity","resonator bifurcation","positive feedback","optical responsivity","critical slowing","frequency-domain multiplexing","quasiparticle noise spectrum"],"falsifier":"Drive a representative TiN or Al MKID to the claimed high-gain point near bifurcation and measure whether the white level of the frequency-noise spectrum and the height of a weak optical calibration tone both rise by the same factor of ~10 while the roll-off frequency falls by a comparable factor; any large excess dissipation or hysteresis that prevents stable operation would falsify the claim.","tokens_in":15146,"feed_emoji":"📡","tokens_out":890,"duration_ms":10225,"temperature":0.7,"pith_summary":"Microwave kinetic inductance detectors are usually run with strong readout tones so amplifier and dielectric noise sit below the signal. Those same tones also change the kinetic inductance through the circulating current, which shifts the resonant frequency. This paper shows that the shift can be turned into a controllable positive-feedback loop: if the tone sits below the driven resonance and power is raised toward the bifurcation point, a small optical load moves the resonance toward the tone, more current flows, and the frequency shift is reinforced. Laboratory maps and noise spectra demonstrate roughly a factor-of-ten gain in optical responsivity together with a matching lengthening of the resonator relaxation time. The authors supply an operating-state map so the strength of the feedback can be dialed in by choice of generator frequency and power. The practical payoff is higher sensitivity whenever the limiting noise is external rather than intrinsic to the detector.","feed_headline":"MKID readout near bifurcation boosts optical gain tenfold","feed_subtitle":"Positive current feedback trades bandwidth for low-frequency responsivity when external noise dominates","key_machinery":"The reactive feedback factor 1 − ∂Ẽ/∂x that appears in the small-signal response of the driven Kerr cavity. When this factor approaches zero (positive-feedback side of the critical point), both optical and generator responsivities are strongly enhanced and the slow eigenvalue of the linearized dynamics vanishes, producing critical slowing.","core_discovery":"When an MKID readout tone is placed below the current-shifted resonant frequency and generator power is tuned near the bifurcation cusp, the reactive readout-current feedback multiplies the optical responsivity by approximately ten relative to the undriven value, while the same feedback lengthens the driven resonator relaxation time and therefore reduces detector bandwidth. The strength of this trade-off is set by the two experimentally controlled coordinates of generator detuning and power.","pith_inferences":["If dissipative nonlinearity remains weak, the same positive-feedback operating region could be used to equalize responsivity across a large array after fabrication scatter.","Critical slowing near the cusp may set a practical upper limit on usable gain for time-variable astronomical signals whose spectrum extends above a few tens of hertz.","Combining active frequency tracking with the nonlinear gain map could allow real-time optimization of the gain-bandwidth product under varying sky loading."],"forward_implications":["MKID bias can be chosen deliberately on an operating-state map to trade bandwidth for low-frequency gain when amplifier or system noise dominates.","Positive-feedback bias supplies an additional lever, beyond raw generator amplitude, for raising signal above fixed additive noise floors.","The same current-dependent frequency shift can be used to keep resonators from colliding under changing optical load, supporting denser frequency multiplexing.","Noise-spectrum shape (white level and roll-off) becomes a direct diagnostic of the chosen feedback strength."],"fun_headline_variants":["MKID readout near bifurcation multiplies optical responsivity by 10","Current feedback yields 10x MKID optical gain while cutting bandwidth","Near-cusp readout power tunes MKID responsivity up tenfold","Nonlinear readout current multiplies MKID optical response by factor of 10","Detuning and power set 10x MKID responsivity-bandwidth trade-off"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The model treats the current nonlinearity as purely reactive and fixes a large reactive-to-dissipative ratio for optical load; if dissipative heating or quasiparticle generation by the readout becomes important near bifurcation, the predicted gain and stability change.","fun_headline_variants_meta":{"raw":{"variants":["MKID readout near bifurcation multiplies optical responsivity by 10","Current feedback yields 10x MKID optical gain while cutting bandwidth","Near-cusp readout power tunes MKID responsivity up tenfold","Nonlinear readout current multiplies MKID optical response by factor of 10","Detuning and power set 10x MKID responsivity-bandwidth trade-off"]},"model":"grok-4.5","effort":"low","cost_usd":0.00337,"raw_usage":{"total_tokens":1111,"prompt_tokens":728,"num_sources_used":0,"completion_tokens":80,"cost_in_usd_ticks":33700000,"prompt_tokens_details":{"text_tokens":728,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":303,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":728,"tokens_out":80,"duration_ms":4447,"temperature":1.0,"reasoning_tokens":303,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-13T04:52:36.867892+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Drive a representative TiN or Al MKID to the claimed high-gain point near bifurcation and measure whether the white level of the frequency-noise spectrum and the height of a weak optical calibration tone both rise by the same factor of ~10 while the roll-off frequency falls by a comparable factor; any large excess dissipation or hysteresis that prevents stable operation would falsify the claim.","supporting_citations":[],"review_version":1}