{"id":"19c8a37c-83eb-4347-9bae-9c870a53ce38","arxiv_id":"2508.19606","paper_version":1,"verdict":"REJECT","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":6,"one_line_summary":"Off-resonant driving of a dissipative Jaynes-Cummings sensor, with field-only homodyne readout and Bayesian estimation, is claimed to nearly saturate the global quantum Fisher information bound.","lead":"This paper studies a driven and dissipative Jaynes-Cummings system as a sensor for the drive amplitude, and claims that working off resonance, and reading out only the field with a homodyne measurement, almost reaches the ultimate precision limit. A smart generalist would read it to see whether practical, partial-access critical sensors can beat standard precision limits.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Abstract's super-linear claim for full system and field is contradicted by the paper's own B≈1 fits (Figs. 2c, 4c); only the low-magnitude qubit QFI has B≈1.25, and the narrow N-range fits lack error bars or convergence checks.","rationale":"The paper's headline claim is a super-linear enhancement in sensitivity with respect to the resource N for both the full system and the partial subsystem. The internal evidence points the other way: the reported fits yield linear scaling for the whole system and the field, and super-linear scaling only for the qubit, whose QFI is orders of magnitude too small to constitute a useful sensing resource. That is an internal inconsistency between the abstract/conclusions and the quantitative results, not merely a disagreement with an external consensus. The narrow fitting window and absence of error bars or convergence checks further weaken the scaling conclusion, but the decisive issue is that the paper's own numbers contradict its advertised message. The homodyne Bayesian estimation result may be plausible, but it does not compensate for an unsupported central claim. I therefore concur with the reader's rejection, though the rejection is based on the scaling claim rather than on any issue with the estimation protocol itself.","tokens_in":18093,"tokens_out":5158,"duration_ms":46314,"concrete_test":"Extract the raw max-QFI data points for the whole system and the field in the off-resonance case (Fig. 4c) and re-fit g^2 Q_j = A N^B + C over N=20-100 and, separately, N=60-200, running new simulations at larger N if needed with explicit truncation-convergence checks. If B for whole and field is indistinguishable from 1 within error bars in both ranges, the abstract's super-linear claim fails; if B is robustly greater than 1 in the extended range, the concern is resolved.","verdict_should_be":"REJECT","load_bearing_attack":"The central advertised result is a super-linear quantum-enhanced sensitivity in both the full system and the partial field subsystem, with N=(g/2κ)^2 as the resource. The paper's own fits do not support this. In 'On resonance scaling analysis', Fig. 2(c) yields B≈1 for the whole system and the field, while Fig. 2(b) gives B≈1.25 only for the qubit. In 'Off resonance scaling analysis' (Fig. 4(c)), the text says the fitted exponent 'can be either linear or super-linear, depending on the specific subsystem' without giving values; the figure labels 'Fit N' and 'Fit N^1.25' are consistent with whole/field being linear and only the qubit super-linear. The abstract and conclusions nevertheless claim the enhancement manifests in both the full system and a partial subsystem. Moreover, the qubit's super-linear QFI is orders of magnitude smaller than the field/whole QFI (Fig. 2b vs 2c), so it does not provide a practical sensing advantage. The fits themselves are fragile: they use only N from about 20 to 62 or 100, no error bars, no truncation-convergence checks, and the three-parameter AN^B+C form can absorb finite-size curvature. The 'near-ultimate homodyne' result is a separate numerical demonstration and does not rescue the scaling claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the driven-dissipative Jaynes-Cummings model as a critical quantum sensor for estimating the drive amplitude E, using the ratio N=(g/2κ)^2 as the sensing resource. It presents numerical QFI calculations for the full system, the field subsystem, and the qubit subsystem, both on and off resonance, and proposes a homodyne measurement combined with Bayesian estimation to approach the ultimate precision bound. The central advertised claim is that detuning off resonance yields a super-linear (quantum-enhanced) scaling of sensitivity with N in both the full system and the field subsystem, while requiring only partial access. The paper also reports that the field subsystem encodes most of the information and that the homodyne protocol nearly saturates the full-system QFI bound.","tokens_in":18542,"tokens_out":4105,"duration_ms":38108,"significance":"If the central scaling claim were correct, the result would be of genuine interest: it would show dissipative critical sensing with no initial-state preparation and partial accessibility, with a concrete feasible measurement. The manuscript contains a systematic numerical study of a physically relevant model, including QFI for the whole system and subsystems, phase-space diagnostics, entanglement and purity analysis, and a self-contained estimation protocol. However, the headline claim of super-linear quantum-enhanced sensitivity is not supported by the reported fits: the full-system and field QFI are fitted with a linear exponent B≈1 in both the on-resonance and the off-resonance analyses, and only the qubit, whose QFI is orders of magnitude smaller, shows a super-linear fit. The homodyne near-saturation result is a separate numerical demonstration and does not remedy the scaling contradiction. As it stands, the advertised main result is contradicted by the manuscript's own evidence.","major_comments":[{"comment":"The abstract and conclusions claim that super-linear enhancement manifests in the full system and a partial subsystem, but Fig. 2(c) reports a linear exponent B≈1 for the whole system and the field subsystem, while only the qubit in Fig. 2(b) has B≈1.25. Since the qubit QFI is about three orders of magnitude smaller than the field/whole QFI (max about 35 in Fig. 2(b) versus about 2×10^4 in Fig. 2(c)), the qubit's super-linear fit does not support the advertised practical quantum-enhanced sensitivity.","section":"'On resonance scaling analysis', Fig. 2"},{"comment":"The text states that the fitted exponent can be linear or super-linear depending on the subsystem, but it does not report the actual B values, fit uncertainties, or goodness-of-fit for any subsystem. The figure labels 'Fit N' and 'Fit N^1.25' indicate that the whole and field subsystems are again linear and only the qubit is super-linear, so the off-resonance data do not substantiate the abstract's claim that super-linear enhancement appears in both the full system and the partial subsystem.","section":"'Off resonance scaling analysis', Fig. 4(c)"},{"comment":"The asymptotic exponents are inferred from power-law fits of the form AN^B+C over a narrow window N≳20 up to N≈62 or 100, with no error bars, no residuals, no fitting-window dependence, and no truncation-convergence checks. Because the three-parameter form can absorb finite-size curvature through the constant C, the reported exponents are not a reliable basis for a claim of super-linear (quantum-enhanced) scaling.","section":"'On resonance scaling analysis' and 'Off resonance scaling analysis', Figs. 2 and 4"},{"comment":"The Bayesian estimation demonstration uses the same theoretical model to generate simulated data (counts C_m drawn from the discretized probabilities P_m) and to compute the likelihood P(data|E). The near-saturation shown in Fig. 5 is therefore a self-consistency check of the model rather than an independent validation that the homodyne protocol reaches the ultimate bound; an independent noise model or experimental data would be needed to support the 'near-ultimate' claim.","section":"Supplemental Material, Sec. II (C-D)"},{"comment":"The off-resonance enhancement relative to resonance is a constant-factor improvement in QFI at fixed N, not a change in scaling with N. With B≈1 for the field and whole system, the QFI grows linearly with the resource N, so the central conclusion that quantum-enhanced sensitivity is achieved for the whole probe and the field subsystem is not supported by the reported data.","section":"'Off resonance quantum-enhanced sensor' and 'Conclusions'"}],"minor_comments":[{"comment":"The y-axis labels 'g^2 whole', 'g^2 field', and 'g^2 qubit' should be written as 'g^2 Q_whole', 'g^2 Q_field', and 'g^2 Q_qubit' to avoid ambiguity.","section":"Fig. 2 and Fig. 4 captions"},{"comment":"The notation in Eq. (6) is nonstandard; the subscript 'n' under arg max appears to be a placeholder for the variable E/g, and the expression should be written explicitly as arg max over E/g of Q_j(E).","section":"Eq. (6)"},{"comment":"Reference [68] is incomplete, lacking a title, and references [17]/[62] and [23]/[57] are duplicated with different numbers.","section":"References"},{"comment":"The text introduces optimal values (Δ/g)_j^* and (E/g)_j^* for each subsystem j but then says the section focuses only on the detuning that maximizes Q_field; the relation between the general definition and the specific values plotted in Fig. 4 should be clarified.","section":"'Off resonance scaling analysis', Eq. (7) and Fig. 4"},{"comment":"The sentence 'Achieving quantum-enhanced sensitivity for the whole probe and the field subsystem' is a sentence fragment and should be merged with the preceding sentence.","section":"Conclusions"}],"recommendation":"reject","confidential_remarks":"The central advertised result is contradicted by the manuscript's own numerical fits, and a revision that retains the current title and abstract would need to replace the super-linear claim with a linear-scaling claim, substantially changing the paper's message. The near-ultimate homodyne result is also weakened by the fact that the simulated data and likelihood are generated from the same model. I do not see a load-bearing fix within the scope of the current manuscript that would make the advertised claims correct."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this paper has a real, useful core — detuning a driven-dissipative Jaynes-Cummings system selects one bistable branch, and local homodyne readout with Bayesian estimation nearly reaches the global QFI bound. That part is internally consistent and worth knowing. But the headline claim that the sensor shows super-linear, quantum-enhanced sensitivity in the full system and field subsystem is not supported by the paper's own numbers.\n\nWhat's actually new: combining off-resonant detuning to pick a preferred phase-space peak, partial field access, and a feasible homodyne estimator. The field does carry nearly all the information, and the near-saturation of the global QFI by a local Gaussian measurement is a nice numerical demonstration. They also avoid ground-state preparation by using the steady state of the dissipative dynamics. Credit where due. The citation pattern is solid, covering the relevant dissipative-phase-transition and partial-access literature.\n\nWhere it wobbles: the scaling analysis. On resonance, Fig. 2(c) fits whole-system and field QFI with B≈1, i.e., linear. Off resonance, Fig. 4(c) says the exponent can be linear or super-linear depending on subsystem, but the abstract and conclusions claim super-linear enhancement in both the full system and a partial subsystem. The only clear super-linear exponent is the qubit's B≈1.25, and its QFI is orders of magnitude smaller than the field's, so it is not practically useful. That mismatch is not a cosmetic detail; it is the central advertised result. The fits themselves are also fragile — N spans roughly 20 to 100, no error bars, no truncation-convergence check, and a three-parameter AN^B+C form can absorb finite-size curvature. The near-saturation part is separate and is not rescued by the scaling claim.\n\nOne smaller thing: the SM uses g/κ=10 for the homodyne angle optimization while the main text says κ=10^{-6}; this may be a units choice, but it should be clarified.\n\nWho this is for: someone working on dissipative critical metrology or partial-access sensing would get value from the homodyne/Bayesian demonstration and the phase-space intuition. As submitted, I would not cite the scaling claim, and the abstract needs to be rewritten to match the data.\n\nRecommendation: send to peer review, but with a clear expectation of major revision. The near-saturation result is substantive enough to warrant referee time, and the scaling issue is fixable if the authors either produce a converged super-linear region for the field/whole system or revise the claims to linear-plus-constant with better characterization. Do not desk-reject; do not accept as is.","headline":"The near-saturation homodyne result is worth a look, but the advertised super-linear scaling does not survive contact with the paper's own fits.","tokens_in":18961,"tokens_out":3035,"would_cite":false,"duration_ms":28362,"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":"Off-resonance operation turns a dissipative Jaynes-Cummings sensor into a near-ultimate partial-access probe.","keywords":["dissipative quantum phase transition","Jaynes-Cummings model","quantum sensing","quantum Fisher information","partial accessibility","homodyne detection","Bayesian estimation","critical quantum metrology"],"falsifier":"Compute the peak quantum Fisher information for $N=100$, $200$, $500$, and $1000$ with the same Liouvillian and refit $A N^B + C$; if $B$ falls to 1 (or below) within the extended range, the claimed super-linear quantum enhancement is not established. A complementary experiment would compare homodyne-based Bayesian estimator variance against $[M g^2 Q_{\\rm whole}(E)]^{-1}$ at a fixed larger $N$, since the paper's near-saturation claim fails if the gap widens as $N$ grows.","tokens_in":17821,"feed_emoji":"⚛️","tokens_out":9698,"duration_ms":87263,"temperature":0.7,"pith_summary":"The paper argues that a driven-dissipative Jaynes-Cummings system—a single qubit coupled to a lossy cavity mode—can serve as a high-precision sensor without any careful preparation of its initial state. Its precision is measured against the resource $N=(g/2\\kappa)^2$, the ratio of coherent qubit-field coupling to photon loss, and the paper claims that detuning the drive off resonance makes the sensor significantly more sensitive than on-resonance operation, with precision growing faster than linearly in $N$. Near the critical point, the qubit and field become only weakly entangled, and nearly all information about the unknown drive amplitude $E$ is carried by the field mode alone. The paper closes the practical gap by showing that homodyne detection of the field, combined with Bayesian estimation, nearly reaches the full-system quantum limit. This matters because it removes two usual requirements—complex initial-state preparation and full-system readout—from critical quantum sensing.","feed_headline":"Detuned sensor nearly hits ultimate precision with partial access","feed_subtitle":"Homodyne detection of the cavity field plus Bayesian estimation nearly saturates the full-system quantum Fisher bound.","key_machinery":"The load-bearing element is the steady state of the driven Jaynes-Cummings master equation, computed as the zero-eigenvalue eigenstate of the Liouvillian. The system's dissipative quantum phase transition at drive amplitude $2E=g$ (in the thermodynamic limit $N=(g/2\\kappa)^2\\to\\infty$) organizes all the sensing behavior: on resonance the field turns from a vacuum-like state to a bistable two-peak Wigner function, and the qubit polarizes so strongly that qubit-field entanglement becomes weak. Detuning $\\Delta$ breaks the bistability and selects one peak, producing a sharp maximum in the quantum Fisher information around $E/g\\approx 0.4$. Because the steady state is weakly entangled and the field's Wigner function stays positive and nearly Gaussian, the field's reduced state carries almost all of the information about $E$, which is why a single optimized field quadrature measurement suffices.","core_discovery":"The paper's central claim is that the steady state of the driven-dissipative Jaynes-Cummings model, taken at its dissipative quantum phase transition, is a complete sensing solution: it needs no initial-state engineering, its optimal operation point is a sharp response peak created by detuning, and its precision can be read out from the field alone. The resource is $N=(g/2\\kappa)^2$, and the maximum quantum Fisher information grows with $N$—with super-linear scaling reported for the qubit subsystem, whose QFI follows an $N^{1.25}$ power law, while the whole-system and field QFIs show the same polynomial growth. Detuning selects one of the two bistable field states in phase space, and at the optimal working point the field's reduced state reproduces almost the whole-system QFI. Combining a homodyne measurement on the field with Bayesian estimation yields estimator variance close to $[M g^2 Q_{\\rm whole}(E)]^{-1}$, the global quantum Cramér-Rao bound, so the practical local scheme nearly saturates the ultimate limit set by the full probe.","pith_inferences":["A testable extension is to port the same homodyne-plus-Bayesian pipeline to other driven-dissipative systems with a single dominant positive phase-space peak, such as detuned Kerr resonators, where the near-Gaussian condition that makes homodyne efficient is also satisfied.","Since off-resonance operation shifts the optimal homodyne angle between $\\pi/2$ and $3\\pi/4$ as $N$ changes, an adaptive protocol that re-optimizes the quadrature angle during Bayesian updating could close the remaining gap to the QFI bound at finite $N$.","If the reported exponents survive at larger $N$, the resource $N$ functions as an effective probe size, and a rigorous finite-size scaling analysis of the maximum QFI would connect this dissipative sensor to the established critical-metrology scaling hierarchy."],"forward_implications":["Steady-state critical sensors of this type can be run without ground-state preparation: the same precision should be reproducible from generic initial conditions because the sensing state is the dissipative steady state.","A field-only homodyne measurement followed by Bayesian updating recovers almost all of the estimation precision available in the full qubit-field state, so the protocol is implementable with linear optics and standard post-processing.","Detuning the drive off resonance and choosing its sign lets an experimenter select the preferred bistable phase-space peak and operate at the sharp response near $E/g\\approx 0.4$, improving on resonance.","Because the field's Wigner function remains positive and near-Gaussian at the working point, nonclassical resource states are not needed for the measurement step, only for the coherent drive and coupling.","Increasing the coupling-to-loss ratio $N=(g/2\\kappa)^2$ directly raises the peak information, giving a concrete design target for cavity-QED or circuit-QED implementations."],"supporting_citations":[{"why":"supplies the driven Jaynes-Cummings model, the resource $N=(g/2\\kappa)^2$, and the dissipative phase transition at $2E=g$ that organizes the critical response.","marker":"[84]"},{"why":"establishes dissipative phase transitions as metrological resources, the baseline this work extends to partial access.","marker":"[74]"},{"why":"introduces the partially accessible many-body sensing setting whose field-only readout strategy this work adapts.","marker":"[8]"},{"why":"provides the quantum Cramér-Rao bound and statistical-distance geometry underlying the QFI-based precision limits.","marker":"[121]"},{"why":"defines classical and quantum Fisher information and the POVM optimization used to state the ultimate bounds.","marker":"[117]"},{"why":"is the numerical toolbox used to obtain steady states as zero eigenvalues of the Liouvillian.","marker":"[122]"}],"fun_headline_variants":["No-prep quantum sensor nears ultimate sensitivity","Partial access still reaches near-ultimate quantum sensing","Field-only readout nearly saturates quantum Fisher bound","Detuned cavity boosts sensitivity without initial-state tuning","Super-linear quantum sensing from a driven cavity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim that precision grows super-linearly with the resource $N$ rests on fitting $A N^B + C$ to numerically computed peak quantum Fisher information over a short range of $N$ (about 20 to a few hundred) with no reported error bars, so if that fitted exponent is a finite-size artifact the super-linear enhancement may not survive at larger $N$.","fun_headline_variants_meta":{"raw":{"variants":["No-prep quantum sensor nears ultimate sensitivity","Partial access still reaches near-ultimate quantum sensing","Field-only readout nearly saturates quantum Fisher bound","Detuned cavity boosts sensitivity without initial-state tuning","Super-linear quantum sensing from a driven cavity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000253,"raw_usage":{"total_tokens":1567,"prompt_tokens":947,"completion_tokens":620,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":563,"completion_tokens_details":{"reasoning_tokens":563}},"tokens_in":563,"tokens_out":620,"duration_ms":6284,"temperature":1.0,"reasoning_tokens":563,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:49:39.426169+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the peak quantum Fisher information for $N=100$, $200$, $500$, and $1000$ with the same Liouvillian and refit $A N^B + C$; if $B$ falls to 1 (or below) within the extended range, the claimed super-linear quantum enhancement is not established. A complementary experiment would compare homodyne-based Bayesian estimator variance against $[M g^2 Q_{\\rm whole}(E)]^{-1}$ at a fixed larger $N$, since the paper's near-saturation claim fails if the gap widens as $N$ grows.","supporting_citations":[{"cited_title":"Nonequilibrium functional renormalization for driven- dissipative bose-einstein condensation,","cited_arxiv_id":null,"evidence_quote":"supplies the driven Jaynes-Cummings model, the resource $N=(g/2\\kappa)^2$, and the dissipative phase transition at $2E=g$ that organizes the critical response."},{"cited_title":"Driven markovian quan- tum criticality,","cited_arxiv_id":null,"evidence_quote":"establishes dissipative phase transitions as metrological resources, the baseline this work extends to partial access."},{"cited_title":"Holevo, Probabilistic and Statistical Aspects of Quantum Theory (Edizioni della Normale, Pisa, 2011)","cited_arxiv_id":null,"evidence_quote":"provides the quantum Cramér-Rao bound and statistical-distance geometry underlying the QFI-based precision limits."},{"cited_title":"Information and the accuracy attainable in the estimation of statistical parameters,","cited_arxiv_id":null,"evidence_quote":"defines classical and quantum Fisher information and the POVM optimization used to state the ultimate bounds."},{"cited_title":"Statistical dis- tance and the geometry of quantum states,","cited_arxiv_id":null,"evidence_quote":"is the numerical toolbox used to obtain steady states as zero eigenvalues of the Liouvillian."}],"review_version":2}