{"id":"6bde505a-028e-483c-9656-03a6487856bb","arxiv_id":"2509.04300","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Frequency-resolved photon counting with a coherently displaced local oscillator can saturate the quantum Fisher information for parameters of a driven two-level emitter, and two-color photon correlations give up to a 10^5 gain for detuning estimation.","lead":"This paper develops a theoretical method to calculate how precisely an atom's properties can be measured by frequency-filtering the light it emits. It shows that choosing the right filter frequencies, linewidths, and a coherent shift can make the measurement optimal, and that two-color photon correlations can provide a large precision boost.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Two-sensor master equation is internally inconsistent: Eq. (4) uses sqrt(εγΓ/2) but Appendix A.2 gives sqrt(εγΓ), so the claimed 10^5 gain is conditional on a factor √2.","rationale":"The reader's weakest-assumption analysis identifies exactly the same load-bearing issue I find: the two-sensor master equation contains a factor-√2 inconsistency between the main-text Eq. (4) and the Appendix A.2 derivation. This concern is load-bearing because the two-sensor quantitative claims—especially the five-orders-of-magnitude gain in F_Δ/F_Δ^I near leapfrog transitions—are the most striking new quantitative results in the paper and depend directly on the cross-coupling coefficient. A factor √2 in the coherent coupling can substantially alter the joint photon-number distribution p(n1,n2|θ) and hence the Fisher information, particularly near points where marginal CFIs are small and the ratio is sensitive. I do not see a reason to escalate the verdict to REJECT: the single-sensor framework, the mean-field-engineering saturation claim, and the general bound of Eq. (19) are independent of this coefficient and appear internally consistent. The missing truncation-order reporting is a secondary reproducibility concern but does not replace the primary coefficient issue. Since the reader already issued a CONDITIONAL verdict and the correct response is to require resolution of this inconsistency before accepting the two-sensor numbers, no verdict change is needed.","tokens_in":31008,"tokens_out":16273,"duration_ms":148159,"concrete_test":"Independently re-derive the two-sensor cascaded master equation for the balanced-beam-splitter setup using the SLH/cascaded input-output formalism, e.g., with collapse operators L1 = sqrt(εγ)σ, L2 = sqrt(Γ)ξ1, L3 = sqrt(Γ)ξ2 and the beam-splitter transformation of Eq. (A7), and check whether the coefficient of ([ξ_i^†, σρ] + [ρσ^†, ξ_i]) is sqrt(εγΓ/2) or sqrt(εγΓ). Then, using whichever coefficient is correct, recompute F_Δ and F_Δ^I at the representative leapfrog point (Δ1, Δ2) = (-2Ω + δω, 2Ω) with Ω = 10γ, Γ = γ, ε = 0.5 (as in Figs. 5-6). If F_Δ/F_Δ^I remains ≈ 10^5, the qualitative claim survives; if it changes by more than an order of magnitude, Figs. 5 and 6 must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing concern is the factor-√2 inconsistency in the two-sensor cascaded master equation. Eq. (4) contains the cross-coupling term -sqrt(εγΓ_i/2){[ξ_i^†, σρ] + [ρσ^†, ξ_i]}, while the derivation in Appendix A.2 leads to Eq. (A12) with -sqrt(Γ1Γ2){...} and the explicit identification sqrt(Γ1Γ2) -> sqrt(εγΓ). For Γ1=γ and Γ2=Γ these two coefficients differ by √2. This is not a cosmetic discrepancy: the source-sensor cross-coupling controls the strength of the frequency-resolved correlations that underpin the headline two-sensor result, F_Δ ≈ 10^5 F_Δ^I near the leapfrog resonance (Figs. 5 and 6). If the appendix is taken literally, all two-sensor Fisher information values, the ratio F_Δ/F_Δ^I, and the optimal linewidth windows in Fig. 6 must be recomputed with the larger coupling. If Eq. (4) is the physically correct equation (as a direct beam-splitter derivation would suggest), then Appendix A.2 has a missing factor 1/√2 in its final step. Either way, the manuscript currently contains two incompatible versions of the central two-sensor model. The single-sensor, mean-field, and Eq. (19) bound results are not affected by this issue.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a cascaded-sensor theoretical framework to quantify the metrological information contained in frequency-filtered photon-counting measurements from a continuously driven two-level emitter. It defines frequency-resolved classical Fisher information (CFI), compares it with the quantum Fisher information (QFI) of the filtered sensor modes, and introduces coherent displacement ('mean-field engineering') as a tunable measurement strategy. In the single-sensor setting, it identifies frequency windows and optimal sensor linewidths, and reports that an optimally chosen displacement can saturate the sensor QFI. It then extends the framework to two sensors coupled to the source through a balanced beam splitter, showing that joint photon counting can exceed independent-detection Fisher information by up to five orders of magnitude near leapfrog resonances. It also proves a general bound, Eq. (19), relating the joint and marginal Fisher informations.","tokens_in":31321,"tokens_out":5880,"duration_ms":61010,"significance":"If the central claims survive scrutiny, the paper makes a useful contribution: it provides a concrete, physically motivated method for benchmarking frequency-resolved measurements in quantum optics, and it identifies a specific measurement strategy—displaced photon counting—that can numerically saturate the QFI of the filtered modes. The proof of Eq. (19) is valuable and independent of the disputed coupling coefficient. The predicted large metrological enhancement from two-photon leapfrog correlations is interesting and falsifiable. The main caveat is that the two-sensor quantitative results rest on a master-equation coupling coefficient that is internally inconsistent; the single-sensor results and the bound in Eq. (19) are not affected.","major_comments":[{"comment":"The two-sensor master equation contains an internal inconsistency. Eq. (4) has cross-coupling coefficient sqrt(ε γ Γ_i / 2), whereas the derivation in Appendix A.2, Eq. (A12), after the stated identifications Γ1→γ, Γ2→Γ, and sqrt(Γ1Γ2)→sqrt(εγΓ), yields coefficient sqrt(εγΓ) — a factor √2 larger. The sentence after Eq. (4) says the 1/√2 accounts for vacuum contributions from the balanced beam splitter, but the beam-splitter input relation (A8) and collective operator (A10) do not introduce a final 1/√2 in the dissipative cross term (A12). Since Eq. (4) determines all two-sensor CFI values, the g^(2) maps, and the headline F_Δ ≈ 10^5 F_Δ^I in Figs. 5 and 6, the two versions cannot both be correct. Please correct one of them. If the corrected coupling differs from Eq. (4), recompute all two-sensor results and state whether the order-of-magnitude gain survives and how the optimal linewidth","section":""},{"comment":"The numerical CFI and QFI require truncating the bosonic sensor Hilbert spaces, but the manuscript does not report the truncation level n_exc or a convergence test. This matters especially for the saturation claim in §III.C: optimal displacements with |α| ~ 4 (Fig. 4) populate many Fock states, so F_α,θ and I_θ must be checked as functions of truncation. Please state the truncation used and demonstrate convergence for the quantities plotted in Figs. 4–6. This is also needed to assess the magnitude of the claimed deviations from QFI saturation near resonance.","section":""}],"minor_comments":[{"comment":"The main text says 'the bound presented in Eq. (18)' but Eq. (18) defines F^I_θ; the actual bound is Eq. (19). Appendix I also says 'Eq. (18)' in its opening sentence. Please correct the cross-reference.","section":""},{"comment":"Typo: 'For the shake of simplicity' should be 'For the sake of simplicity'.","section":""},{"comment":"The range 'Γ ∈ (10^-2, 10)Ω' is ambiguous; it should be written as Γ ∈ (10^{-2}Ω, 10Ω) or similar.","section":""},{"comment":"The parameter value 'Ω = γ/2√γ' appears dimensionally inconsistent; it should be made consistent with the text (e.g., Ω = γ/(2√2) or a stated value in units of γ).","section":""},{"comment":"There is a duplicated and slightly garbled sentence: 'This behaviour is similar to what occurs for θ=γ, except...' appears twice with different wording. Please revise.","section":""}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope and the general approach is sound, but the two-sensor quantitative claims are conditional on resolving the factor √2 discrepancy between Eq. (4) and Appendix A.2. I recommend asking the authors to reconcile the derivation and recompute the affected results; the single-sensor framework and the Eq. (19) bound can stand independently. Also request numerical truncation/convergence details before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Martin,\n\nShort version: the single-sensor and mean-field results are solid and worth building on, but the two-sensor story has an internal inconsistency that needs sorting before the 10^5 gain claim can be taken at face value.\n\nWhat's new and good: the paper turns frequency filtering, a routine tool, into a systematic metrological resource. The cascaded-sensor model for filtered modes is well known, but combining it with Fisher information and coherent displacement is new. The single-sensor analysis is careful: the CFI is computed from the full photon-counting distribution, the linewidth optimization is physically motivated, and the mean-field engineering result—that an optimal displacement lets the CFI saturate the sensor QFI—is a clean and useful observation. The bound in Eq. (19) is also a nice, non-obvious addition, and it's proved correctly in Appendix I. I have no issue with the central framework or the single-sensor numerics.\n\nThe soft spot is the two-sensor master equation. Eq. (4) uses a cross-coupling sqrt(εγΓ/2), which is the physically correct coefficient for a beam-splitter split: each arm gets half the intensity, so the amplitude is reduced by 1/√2. But the derivation in Appendix A.2 ends with sqrt(Γ1Γ2) in Eq. (A12), and the identification sqrt(Γ1Γ2) → sqrt(εγΓ) gives sqrt(εγΓ), not sqrt(εγΓ/2). So the appendix contradicts the main text by a factor of √2. This is not cosmetic: it controls the strength of the frequency-resolved correlations that produce the two-sensor gain. If Eq. (4) is right—which I believe it is, given the explicit beam-splitter argument in the main text—then Appendix A.2 has a missing 1/√2 in its final step, and the paper needs a correction. If the appendix were right, all the two-sensor Fisher information numbers would need recomputing. As written, the manuscript contains two incompatible versions of the model, and a referee cannot tell which numbers to trust.\n\nMinor issues: the numerical truncation order and convergence checks are not reported, and no code is provided. These are not fatal but do hamper reproducibility.\n\nWho is this for? Quantum metrology people interested in continuous measurement and frequency-resolved detection, and quantum optics folks who care about multiphoton correlations as a resource. It deserves a serious referee, but the referee should be told to focus on the coupling coefficient discrepancy and to extract a clear statement of which equation is correct. I'd accept for peer review with the expectation of a moderate revision: fix the appendix, state the truncation, and ideally release the code or at least a clear numerical recipe.\n\nMy verdict: conditional, leaning positive. The framework is real, the single-sensor results are credible, and the two-sensor claim is probably right in spirit but needs the algebra fixed before it becomes citable.\n\nCheers,\n[Your name]","headline":"A genuinely new framework for frequency-resolved metrology, with a fixable but load-bearing inconsistency in the two-sensor appendix that must be resolved before the headline gain can be trusted.","tokens_in":31833,"tokens_out":3076,"would_cite":true,"duration_ms":29319,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81V80","81P50"],"pacs":["42.50.-p","03.65.Ta"],"model":"deepseek-v4-flash","headline":"Frequency-resolved photon counting can be made optimal for sensing emitter parameters by combining spectral filters with a coherent displacement of the detected field.","keywords":["quantum metrology","frequency-resolved photon counting","Fisher information","resonance fluorescence","cascaded quantum systems","mean-field engineering","Mollow triplet","quantum sensing"],"falsifier":"Recompute the two-sensor classical Fisher information at the leapfrog point (Delta1, Delta2) approximately (-2 Omega + delta_omega, 2 Omega) using the cross-coupling coefficient sqrt(epsilon gamma Gamma) instead of sqrt(epsilon gamma Gamma / 2), and check whether F_Delta / F_Delta^I remains near 10^5. An experimental alternative: measure the joint photon-counting distribution of resonance fluorescence at those two frequencies with a tunable filter linewidth and compare the empirical Fisher information with the two predictions.","tokens_in":30871,"feed_emoji":"⚛️","tokens_out":5904,"duration_ms":55562,"temperature":0.7,"pith_summary":"This paper sets out to quantify how much metrological information can be extracted from the frequency-resolved light emitted by a continuously driven quantum system. Modeling spectral filters as cascaded bosonic sensors, the authors compute the full photon-counting statistics of the filtered modes and the associated classical Fisher information, and show that both filter frequency and linewidth strongly control estimation precision, with an optimal linewidth window roughly between 10^-2 times and 10 times the Rabi frequency. They find that coherently displacing the detected field before photon counting—mean-field engineering—can push the classical Fisher information up to the quantum Fisher information of the sensor modes, effectively realizing the optimal measurement. In a two-sensor setup, retaining photon-photon cross-correlations yields a metrological gain that can reach five orders of magnitude near Mollow leapfrog transitions when estimating the qubit-laser detuning, and the paper proves a general bound showing joint detection always gives at least half the sum of the marginal Fisher informations. The relevance is that spectral filtering, already ubiquitous in experiments, can be turned into a principled quantum sensing tool.","feed_headline":"Spectral photon counting hits the quantum metrology limit","feed_subtitle":"With a coherent displacement, frequency-resolved counting extracts the full Fisher information of the filtered light.","key_machinery":"The central object is the cascaded-sensor master equation: each frequency filter is a bosonic mode with detuning Delta_xi and linewidth Gamma, driven by the emitter's output field through a cascaded coupling, so that the filtered radiation is described by a finite-dimensional density matrix. The workhorse quantities are the diagonal elements of that density matrix—the photon-counting probabilities—whose parameter derivatives define the classical Fisher information, and the coherent displacement operator D(alpha) that implements mean-field engineering by mixing the signal with a local oscillator before counting. This machinery converts a measurement-theoretic question (what can be learned fro","core_discovery":"The central claim is that frequency-resolved photon counting on light emitted by a coherently driven two-level emitter can be made optimal for parameter estimation, provided the filtered modes are modeled correctly and the measurement is chosen well. The authors model each filter as a bosonic sensor in a cascaded master equation, reconstruct its steady-state density matrix and its derivative, and use these to evaluate the classical Fisher information of the photon-number distribution. They show that the single-sensor classical Fisher information displays a Mollow-triplet-like spectrum with additional features tied to higher-order photon correlations, that a sensor linewidth in the window Gam","pith_inferences":["The same framework should apply to other continuously driven sources with richer spectra, where the optimal filtering strategies will likely differ; the paper's qualitative conclusion—correlations help for some parameters and not others—implies that parameter-specific optimization is essential.","The factor sqrt(2) discrepancy between the two-sensor master equation in Eq. (4) and its derivation in Appendix A.2 is a load-bearing detail: if the appendix is correct, all quantitative two-sensor results, including the claimed 10^5 gain and the optimal linewidth windows, need to be recomputed, although the qualitative mechanism may survive.","One could test the general bound experimentally by measuring the joint and marginal photon-counting distributions of a resonance-fluorescence source and comparing their Fisher informations; the bound predicts the joint information is never less than half the sum of the marginals.","Mean-field engineering could be combined with spectral filtering in a single setup, potentially allowing parameter estimation at the sensor quantum Fisher information while retaining frequency selectivity."],"forward_implications":["For any driven emitter described by a Lindblad master equation, the same construction yields the optimal filter frequency and linewidth for estimating a given parameter, making spectral filtering a systematic metrological resource rather than an ad hoc tool.","Near the optimal linewidth window, the single-sensor classical Fisher information is enhanced by over ten orders of magnitude relative to the narrowband and broadband limits, so filter bandwidth is a first-order control knob for sensitivity.","An experimentalist who can add a tunable local oscillator to a spectral filter can, in principle, reach the quantum Fisher information of the filtered mode without designing a more complex POVM.","Joint detection on two frequency channels can beat independent measurements by up to five orders of magnitude when the parameter is the detuning and the channels sit on a Mollow leapfrog transition; this makes Hanbury-Brown-Twiss-style setups useful for metrology, not just for correlation measurements.","The inequality F_theta[joint] >= 1/2(F_theta[marginal1] + F_theta[marginal2]) holds generally, so joint measurements can never lose more than a factor of two, but also guarantees no universal advantage of correlations."],"supporting_citations":[{"why":"Supplies the cascaded-sensor method for frequency-filtered and time-resolved photon correlations that the paper uses to reconstruct the filtered density matrix.","marker":"[42]"},{"why":"Establishes the cascaded master equation for a sensor driven by quantum light, the basis of the single-sensor equation.","marker":"[43]"},{"why":"Extends the cascaded description to a two-level source, grounding the source-sensor coupling used in Eq. (3).","marker":"[44]"},{"why":"Provides the cascaded input-output formalism and quantum Ito calculus underlying the Appendix A derivations.","marker":"[60]"},{"why":"Identifies the leapfrog two-photon transitions in the Mollow ladder and their frequency conditions, which the two-sensor gain is tied to.","marker":"[49]"},{"why":"Gives the coherent-field interference method for tuning photon statistics, which the paper calls mean-field engineering.","marker":"[71]"},{"why":"Demonstrates mean-field engineering experimentally, motivating the displacement protocol used to saturate the quantum Fisher information.","marker":"[72]"},{"why":"Supplies the quantum Fisher information bound for continuously measured radiation and the observation that optimal sensitivity occurs away from resonance.","marker":"[34]"},{"why":"Previous work on entanglement of filtered modes from the same source; the paper contrasts its waveguide post-processing master equation with the beam-splitter setup of Eq. (4).","marker":"[40]"}],"fun_headline_variants":["Spectral counting becomes optimal with coherent displacement","Cascaded filter model reveals optimal spectral metrology","Frequency-resolved photon counting reaches Fisher bound","Quantum sensing optimized via spectral filter design","Filtered light metrology: optimal counting strategy found"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that the cascaded coupling coefficient in the two-sensor master equation is sqrt(epsilon gamma Gamma / 2), as written in Eq. (4); the appendix derivation instead gives sqrt(epsilon gamma Gamma), and if the appendix is right, every two-sensor Fisher information value, including the reported 10^5 gain, changes.","fun_headline_variants_meta":{"raw":{"variants":["Spectral counting becomes optimal with coherent displacement","Cascaded filter model reveals optimal spectral metrology","Frequency-resolved photon counting reaches Fisher bound","Quantum sensing optimized via spectral filter design","Filtered light metrology: optimal counting strategy found"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000362,"raw_usage":{"total_tokens":1761,"prompt_tokens":686,"completion_tokens":1075,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":430,"completion_tokens_details":{"reasoning_tokens":1006}},"tokens_in":430,"tokens_out":1075,"duration_ms":10490,"temperature":1.0,"reasoning_tokens":1006,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T10:13:11.076439+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the two-sensor classical Fisher information at the leapfrog point (Delta1, Delta2) approximately (-2 Omega + delta_omega, 2 Omega) using the cross-coupling coefficient sqrt(epsilon gamma Gamma) instead of sqrt(epsilon gamma Gamma / 2), and check whether F_Delta / F_Delta^I remains near 10^5. An experimental alternative: measure the joint photon-counting distribution of resonance fluorescence at those two frequencies with a tunable filter linewidth and compare the empirical Fisher information with the two predictions.","supporting_citations":[{"cited_title":"Cabot, F","cited_arxiv_id":null,"evidence_quote":"Supplies the cascaded-sensor method for frequency-filtered and time-resolved photon correlations that the paper uses to reconstruct the filtered density matrix."},{"cited_title":"Alushi, A","cited_arxiv_id":null,"evidence_quote":"Establishes the cascaded master equation for a sensor driven by quantum light, the basis of the single-sensor equation."},{"cited_title":"Silva, C","cited_arxiv_id":null,"evidence_quote":"Provides the cascaded input-output formalism and quantum Ito calculus underlying the Appendix A derivations."},{"cited_title":"Godley and M","cited_arxiv_id":null,"evidence_quote":"Identifies the leapfrog two-photon transitions in the Mollow ladder and their frequency conditions, which the two-sensor gain is tied to."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the coherent-field interference method for tuning photon statistics, which the paper calls mean-field engineering."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates mean-field engineering experimentally, motivating the displacement protocol used to saturate the quantum Fisher information."},{"cited_title":"Montenegro, M","cited_arxiv_id":null,"evidence_quote":"Previous work on entanglement of filtered modes from the same source; the paper contrasts its waveguide post-processing master equation with the beam-splitter setup of Eq. (4)."}],"review_version":1}