{"id":"1a004e70-a69c-4cf7-aea2-0d980ffc0f71","arxiv_id":"2412.15061","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Squeezing an atomic ensemble both before and after a phase signal deamplifies the encoded phase and nearly doubles the unambiguous measurement range while staying near the optimal quantum limit.","lead":"A new 'quantum deamplification' protocol squeezes atoms both before and after a phase signal, letting a quantum sensor measure over nearly a full period of phase while staying close to the optimal quantum limit. It offers atomic clocks and other precision sensors a simpler, noise-robust route to wide dynamic range than current multi-ensemble or deep-circuit designs.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the linear-estimator and OQI-comparison concerns do not threaten the central QD claim.","rationale":"The reader's weakest-assumption choice (linear estimator) is a legitimate numerical caveat, but it is not a threat to the central claim because any optimal nonlinear estimator can only reduce the BMSE for a fixed measurement; the reported values are conservative upper bounds. I partially agree with the reader that the numerical support could be strengthened by uncertainty quantification and an optimal-estimator check, and that the supplementary OQI discussion needs careful reading. The central physical mechanism, the near-OQI performance at the demonstrated sizes, the decoherence analysis, and the hybrid-sensing robustness all hold up under scrutiny. The most valuable single verification would be an N-scaling study with an optimal estimator, because the paper claims scalability but shows only N=100 in the main text and N=64 in the supplementary. Since no internal inconsistency or clear error was found, the reader's CONDITIONAL verdict should stand unchanged.","tokens_in":17261,"tokens_out":21071,"duration_ms":209777,"concrete_test":"Recompute the QD BMSE at the Fig. 2(b) optimum using the full Bayesian posterior-mean estimator instead of the linear estimator am, for N=100 and N=400, and compare the gap to the exact finite-N OQI bound. If the gap does not grow with N and the nonlinear-estimator BMSE is no larger than the reported linear-estimator value, the central claim is confirmed and the linear-estimator and scalability concerns are settled.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I read the paper in good faith and cannot identify a flaw that would break the central claim. The QD-vs-OQI comparison is a numerical demonstration that a two-segment TACT circuit can approach the optimal single-encoding Bayesian bound, and the core mechanism (squeezing-encoding-squeezing) is supported by the squeezing-parameter and MSE plots. The reader's linear-estimator concern is not load-bearing in the direction implied: for a fixed POVM the Bayesian posterior-mean estimator minimizes the BMSE, so the reported linear-estimator BMSE is an upper bound on the achievable error; replacing it with a nonlinear estimator could only lower the QD curve, making the 'approaches OQI' statement stronger, not weaker. The supplementary statement that sequential QD can beat the OQI at large δphi can be read consistently: OQI is the one-time-encoding bound, while the convergence statement refers to the global minimum of delta-phi/delta-phi over the prior width, so the two are not necessarily contradictory. The main residual risk is that the headline near-OQI result is shown only for N=100 (and N=64 in the supplementary), with no explicit N-scaling study, and the numerical optimization is not accompanied by shipped code. These are completeness gaps rather than demonstrated errors, so they do not change the conditional verdict.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes a 'quantum deamplification' (QD) mechanism to extend the dynamic range of phase estimation while maintaining near-optimal sensitivity. The protocol uses two sequential TACT squeezing operations, one before and one after phase encoding, with the second operation deamplifying the encoded signal. The authors numerically optimize the two TACT durations and a linear estimator slope to minimize the Bayesian mean squared error (BMSE) under a Gaussian prior. They report that the QD sensor approaches the optimal quantum interferometer (OQI) limit, that sequential QD interspersed with multiple phase encodings extends the dynamic range further, and that a hybrid QD+QA two-ensemble scheme improves robustness to detection noise. The numerical results use N=100 in the main text and N=64 in the supplementary material.","tokens_in":17549,"tokens_out":10827,"duration_ms":99644,"significance":"If correct, the QD protocol offers a simple two-TACT-circuit route to near-optimal Bayesian phase estimation with nearly full-period dynamic range, which is directly relevant to atomic clocks and entanglement-enhanced interferometry. The numerical work is internally consistent within the main text, and the data are openly deposited at Zenodo, which is a concrete reproducibility strength. The paper also provides a useful mechanistic picture through Wineland-parameter evolution and Wigner-function comparisons. The main weaknesses are the lack of an N-scaling study, the ambiguous resource accounting in the hybrid sensing comparison, and an apparent inconsistency in the supplementary discussion of sequential QD versus the OQI limit; these need to be clarified before publication.","major_comments":[{"comment":"The supplementary states that sequential QD schemes 'are capable of obtaining even smaller BMSE than the OQI limit at large δφ', but then immediately concludes that 'the optimal performance of sequential QD can only converge to the OQI rather than beating the limit.' These statements are contradictory. Since the main text (Fig. 3(d)) claims convergence to the OQI with increasing n, please specify the resource constraints under which each statement holds. In particular, if sequential QD uses n phase-encoding segments, the comparison should be against an OQI with the same number of encodings, not the one-encoding OQI shown in Fig. 2.","section":"Supplementary Sec. II and Fig. S3(a)"},{"comment":"The BMSE is minimized over t1 and t2 while fixing the linear estimator φ_est(m)=a m. For any fixed POVM, the posterior-mean estimator minimizes the BMSE, so the reported QD curve is an upper bound on the achievable error. This direction is conservative for the claim that QD approaches the OQI. However, the optimal values of t1 and t2 for the posterior-mean estimator may differ from the values reported here, and the size of the gap is not quantified. Please compare the linear-estimator BMSE with the posterior-mean BMSE at representative δφ values, or state explicitly that the linear estimator is a deliberate suboptimal choice whose replacement would only reduce the gap to the OQI.","section":"Section 'QD-based sensing with TACT interactions' and Fig. 2(b-c)"},{"comment":"The resource accounting for the hybrid sensor is unclear. The text says that 'the same total number of particles' is used in all schemes, but the parenthetical 'either split into two independent sensors of N atoms each or measured independently twice with N atoms' appears to assign N atoms to each sensor in the multi-sensor cases, which would give 2N total atoms. Please state the atom number per sensor and the total atom number for each curve in Fig. 4(c-d). Without this clarification, the claim that the hybrid sensor 'consistently outperforms' the other schemes cannot be fully evaluated.","section":"Section 'Hybrid sensing with adaptive measurement' and Fig. 4"},{"comment":"The abstract describes the protocol as 'scalable,' but the numerical demonstrations are restricted to N=100 in the main text and N=64 in the supplementary, with no N-scaling study. Please add at least one additional value of N or derive the expected N dependence of the QD-to-OQI gap, so that the scalability claim is supported by the presented evidence.","section":"Abstract and Conclusion"}],"minor_comments":[{"comment":"The sentence 'Finally, we comment that the sequence QD BMSE through extending interrogation time' is incomplete and should be rewritten to state the intended point about the role of extended interrogation time.","section":"Supplementary Sec. II"},{"comment":"The phrase 'highlighted by the shading area' is not defined; please specify in the caption which curve or region the shading refers to.","section":"Fig. 4(c)"},{"comment":"The parenthetical 'either split into two independent sensors of N atoms each or measured independently twice with N atoms' should be corrected to 'N/2 atoms each' if equal total resources are intended, and the total atom number should be stated explicitly.","section":"Section 'Hybrid sensing with adaptive measurement'"},{"comment":"The Zenodo deposit is a useful step, but the paper would be more reproducible if the code used to generate Figs. 2-4 were also deposited, since all central figures are numerical.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The central QD mechanism is sound and the numerical demonstration is credible. The main revision points are the supplementary's sequential-QD/OQI contradiction and the ambiguous resource accounting in the hybrid scheme. These are fixable within the scope of the paper, and I do not see a reason to reject."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: a solid protocol paper with a simple, convincing central mechanism. Two TACT squeezes around the phase encoding—squeezing-encoding-squeezing—deamplify the phase instead of amplifying it, and the numerics show a near-OQI Bayesian mean squared error over the full [−π,π] range. I think the core claim holds.\n\nWhat's actually new is the framing of this as quantum deamplification and the explicit benchmark against the OQI limit. The squeezing-encoding-squeezing sequence is close to the authors' earlier cyclic nonlinear interferometry, but the dynamic-range extension and the two-gate near-optimality are fresh, and the hybrid QD+QA scheme is a sensible practical addition. The citations to their own prior work are appropriate; they're not hiding the connection.\n\nThe central mechanism is physically transparent: the second squeezing compresses the encoded phase into the classical unambiguous window, so the dynamic range grows at modest cost in local sensitivity. The Wigner-function comparison in the supplement gives intuition for why the QD sensor tracks the OQI.\n\nMain soft spots. First, the supplement says sequential QD can beat the OQI limit at large δϕ, then a few sentences later says it can only converge to the OQI. That is a real internal contradiction and needs fixing. Second, the numerics are shown only at N=100 (main text) and N=64 (supplement); there is no scaling study and no uncertainty quantification on the optimization. The data are available on Zenodo, but no code is shipped, which makes it harder to audit the 'nearly approaches OQI' claim. I don't think these are fatal, but they're more than cosmetic. The linear-estimator concern the reader raised is not a problem: for fixed t1,t2 a linear estimator is an upper bound on Bayesian MSE, so switching to a nonlinear estimator could only lower the QD curves, which would make the paper's claim stronger, not weaker.\n\nWho this is for: anyone building entanglement-enhanced clocks or Ramsey interferometers with TACT/OAT or cavity-QED platforms. It deserves a serious referee. My recommendation: send it to peer review. Ask the authors to reconcile the supplementary contradiction, add a couple of N values or a scaling statement, and consider releasing code or at least a detailed pseudocode for the optimization.","headline":"A simple squeezing-encoding-squeezing protocol that likely extends dynamic range to nearly full period while approaching the OQI; solid and worth refereeing, despite a small internal contradiction in the supplement and thin numerical scope.","tokens_in":18097,"tokens_out":3009,"would_cite":true,"duration_ms":20536,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A squeeze-encode-squeeze sequence can deamplify the encoded phase, letting a single ensemble stay near the optimal Bayesian quantum limit over almost the full phase period.","keywords":["quantum metrology","spin squeezing","two-axis counter-twisting","dynamic range","Bayesian phase estimation","optimal quantum interferometer","detection noise","atomic clocks"],"falsifier":"Take the same optimized QD states and compute the Bayesian mean squared error with an optimal estimator, such as the posterior mean or a maximum-likelihood grid over $\\phi$, instead of the fitted linear slope $\\phi_{\\mathrm{est}}(m)=am$; if the resulting errors at large $\\delta\\phi$ differ noticeably from the reported curves, or if an experimental measurement of $\\hat{S}_y$ for phases near $\\pm\\pi$ shows the error rising sharply, the claim that QD nearly approaches the OQI limit over the full range would be refuted.","tokens_in":17063,"feed_emoji":"⚛️","tokens_out":9969,"duration_ms":82415,"temperature":0.7,"pith_summary":"The paper argues that the usual quantum-amplification trick can be inverted: instead of amplifying a small phase to beat noise, one can squeeze again after the phase is encoded, deamplifying the phase signal so it stays inside the unambiguous interval of the readout. Using two-axis counter-twisting (TACT) interactions, this squeeze-encode-squeeze sequence turns a single ensemble into a Bayesian phase sensor whose error stays low across nearly the full $[-\\pi,\\pi]$ range, approaching the optimal quantum interferometer (OQI) limit. Adding more squeeze-encode-squeeze layers extends the unambiguous range beyond $[-\\pi,\\pi]$ and lowers the Bayesian error further, and a hybrid scheme that pairs deamplification with amplification keeps the protocol accurate when detection is noisy. If this works as claimed, entanglement-enhanced clocks and interferometers could gain a much wider operating range without giving up much sensitivity.","feed_headline":"Double squeezing widens quantum phase sensors to nearly ±π","feed_subtitle":"A second two-axis twist deamplifies the phase, holding Bayesian error near the best single-encoding quantum limit across the whole range.","key_machinery":"The load-bearing object is quantum deamplification (QD): a squeezing-encoding-squeezing sequence in which the second spin-squeezing stage is applied with the same sign as the first, so it shrinks the phase-dependent signal instead of undoing the squeezing. Concretely, the paper uses the two-axis counter-twisting (TACT) Hamiltonian $\\hat{H}_{\\mathrm{TACT}} = -\\chi(\\hat{S}_y\\hat{S}_z+\\hat{S}_z\\hat{S}_y)$, with $t_1$ for probe preparation and $t_2$ for interaction-based readout, and quantifies the intermediate squeezing by the Wineland parameter $\\xi_W^2=N \\Delta^2 \\hat{S}_{\\perp,\\min}/|\\langle\\hat{S}\\rangle|^2$. The same QD mechanism is generalized in the supplementary material to other collective-spin models, and the hybrid protocol combines QD with quantum amplification (QA), where the second segment has the opposite sign.","core_discovery":"The central claim is that two sequential TACT squeezing segments, one before and one after phase encoding, act as a quantum deamplifier: the second segment compresses the phase-dependent part of the state, mapping the encoded phase $\\phi$ to a smaller deamplified phase $\\varphi$ that lies within the classical ambiguity-free window. The paper demonstrates numerically that, with a linear estimator $\\phi_{\\mathrm{est}}(m)=am$ and a Gaussian prior, there are squeezing times $t_1,t_2$ such that the Bayesian mean squared error of the QD sensor nearly reaches the OQI limit, and the error is reduced throughout the whole $[-\\pi,\\pi]$ interval rather than only near zero. Sequential QD, with several phase-encoding layers interleaved with squeezing, extends the unambiguous range beyond $[-\\pi,\\pi]$ and further decreases the BMSE, converging toward the OQI limit as the number of layers grows. Finally, a two-sensor hybrid design, in which a QD sensor provides a coarse wide-range estimate and a QA sensor provides a fine corrected estimate, keeps the Bayesian error low and degrades slowly as detection noise strength $\\sigma_{\\mathrm{det}}$ increases.","pith_inferences":["Beyond the paper: an optimal nonlinear estimator could be tested against the fitted linear slope; if the gap to the OQI widens or closes significantly, the 'minimal cost of sensitivity' conclusion would need to be re-stated in terms of the estimator used.","Beyond the paper: the sequential-QD result that BMSE can dip below the one-time-encoding OQI curve at large $\\delta\\phi$ likely reflects the OQI's single-encoding constraint rather than a real violation; a fairer benchmark would allow the OQI multiple encoding passes or compare at equal total interaction time.","Beyond the paper: the hybrid QD+QA feedback protocol could be combined with multi-ensemble averaging to reduce quantum projection noise while keeping the bias error from phase slips low, a combination the paper treats only separately."],"forward_implications":["A single ensemble, with two TACT squeezes, can simultaneously give near-optimal Bayesian precision and a nearly $[-\\pi,\\pi]$ unambiguous range, removing the need for multiple ensembles or auxiliary interferometers in that operating regime.","For atomic clocks, the wider unambiguous window allows longer phase interrogation times without phase-slip errors, which directly targets long-term frequency stability.","Sequential QD pushes the unambiguous range past $[-\\pi,\\pi]$, at the price of some local sensitivity, and its BMSE approaches the OQI limit as the number of layers grows.","Pairing QD with QA in the hybrid two-sensor design preserves most of the wide-range advantage while resisting detection noise better than QD alone, and the same deamplification idea transfers to other collective-spin models."],"supporting_citations":[{"why":"Introduces spin squeezing and the TACT Hamiltonian that defines the model used throughout.","marker":"[4]"},{"why":"Establishes the variational Ramsey-interferometer framework and the Bayesian-optimization baseline the QD protocol is tested against.","marker":"[30]"},{"why":"Supplies the idea of interaction-based readout that motivates the second squeezing segment.","marker":"[37]"},{"why":"Provides the detection-noise model and robustness results the paper adapts for its noise analysis.","marker":"[38]"},{"why":"Demonstrates time-reversal quantum amplification, the contrast case whose limited dynamic range QD improves on.","marker":"[45]"},{"why":"Source of the linear-estimator and Bayesian-mean-squared-error procedure used in the numerical study.","marker":"[60]"},{"why":"Experimental demonstration of optimized Ramsey sensing that QD aims to simplify or match.","marker":"[61]"},{"why":"Defines the optimal quantum interferometer limit that QD-based sensing is claimed to approach.","marker":"[62]"},{"why":"Shows cavity-QED engineering of TACT dynamics, supporting experimental feasibility.","marker":"[63]"},{"why":"Demonstrates two-axis twisting with polar molecules, supporting experimental feasibility.","marker":"[64]"}],"fun_headline_variants":["Quantum deamplification stretches phase sensing to near full circle","Two twists widen quantum sensor range to almost ±π","Sequential squeezing extends quantum metrology dynamic range","Deamplified quantum sensing achieves near-optimal precision over full range"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The numerical case rests on recovering the phase with a straight-line estimator $\\phi_{\\mathrm{est}}(m)=am$ whose slope is fitted to the measurement distribution; the paper does not prove this estimator is near-optimal for its highly non-Gaussian final states, and the reported Bayesian errors also assume a Gaussian prior centered at zero and a fixed atom number.","fun_headline_variants_meta":{"raw":{"variants":["Quantum deamplification stretches phase sensing to near full circle","Two twists widen quantum sensor range to almost ±π","Sequential squeezing extends quantum metrology dynamic range","Deamplified quantum sensing achieves near-optimal precision over full range"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000872,"raw_usage":{"total_tokens":3789,"prompt_tokens":971,"completion_tokens":2818,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":2752}},"tokens_in":587,"tokens_out":2818,"duration_ms":17243,"temperature":1.0,"reasoning_tokens":2752,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:40:06.029344+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the same optimized QD states and compute the Bayesian mean squared error with an optimal estimator, such as the posterior mean or a maximum-likelihood grid over $\\phi$, instead of the fitted linear slope $\\phi_{\\mathrm{est}}(m)=am$; if the resulting errors at large $\\delta\\phi$ differ noticeably from the reported curves, or if an experimental measurement of $\\hat{S}_y$ for phases near $\\pm\\pi$ shows the error rising sharply, the claim that QD nearly approaches the OQI limit over the full range would be refuted.","supporting_citations":[{"cited_title":"Kaubruegger, D","cited_arxiv_id":null,"evidence_quote":"Establishes the variational Ramsey-interferometer framework and the Bayesian-optimization baseline the QD protocol is tested against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the idea of interaction-based readout that motivates the second squeezing segment."},{"cited_title":"Colombo, E","cited_arxiv_id":null,"evidence_quote":"Demonstrates time-reversal quantum amplification, the contrast case whose limited dynamic range QD improves on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the linear-estimator and Bayesian-mean-squared-error procedure used in the numerical study."},{"cited_title":"Macieszczak, M","cited_arxiv_id":null,"evidence_quote":"Defines the optimal quantum interferometer limit that QD-based sensing is claimed to approach."},{"cited_title":"Miller, A","cited_arxiv_id":null,"evidence_quote":"Demonstrates two-axis twisting with polar molecules, supporting experimental feasibility."}],"review_version":1}