{"id":"fc9ce048-5b1e-430a-b279-f3b44709e2ce","arxiv_id":"2411.14929","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Two photonic models with antiunitary symmetry saturate the multi-parameter quantum Cramér-Rao bound; the mutually conjugate model halves the weighted mean squared error compared with the parallel-copy benchmark.","lead":"An optical experiment shows that two quantum states encoded with mutually conjugate phases can be measured together to reach the fundamental precision limit for estimating two parameters at once. The approach removes the usual trade-off between parameters and gives about twice better precision than identical-copy encoding.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The theoretical QCRB saturation is sound, but the experimental demonstration lacks raw data and uncertainty quantification, leaving the central experimental claim unverified.","rationale":"The reader's verdict of CONDITIONAL is appropriate, but the specific weakest assumption identified by the reader (resource cost of U*) is not the most load-bearing concern for the qubit models used here. For any SU(2) qubit unitary, U* is obtained by conjugating U with the fixed Pauli-Y gate, so the MCM can be implemented with the same number of U calls as the parallel model plus fixed local gates; the paper's own citations [56-59] support exact qubit time-reversal. The theoretical derivation checks out: the proposed POVMs achieve F = Q, so the QCRB saturation and the factor-of-two comparison to the parallel model's HCRB are valid. The genuine weakness is experimental: no raw data, no error bars, no fidelity calibration, and no code are provided, so the central experimental demonstration cannot be independently verified. This supports keeping the verdict CONDITIONAL rather than ACCEPT, while not rising to REJECT because the theoretical argument is coherent and the experimental methods are plausible. The agreement_with_reader is 'partial' because the verdict aligns but the chosen load-bearing concern differs.","tokens_in":9399,"tokens_out":48942,"duration_ms":419827,"concrete_test":"Request the raw photon-count data for the MCM runs (theta = 0.4, 1.0, 1.4 rad; 10 phi values each), recompute the Bayesian posterior WMSE from these counts, and bootstrap the WMSE estimates to attach error bars. Also perform quantum state tomography on the polarization-path two-qubit MCM state to extract preparation fidelity. If the bootstrapped WMSE values overlap the QCRB lines within 2 sigma and the state fidelities exceed ~0.99, the experimental claim is supported; otherwise the demonstration is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The analytical core of the paper is internally consistent: for the MCM state |psi>|psi*>, the Bell POVM of Eq. (10) yields a classical Fisher information matrix equal to the doubled QFIM, F = 2 diag(1, sin^2 theta), so the QCRB is saturated with no trade-off, and the factor-of-two advantage over the parallel model's HCRB follows. The reader's concern about the resource cost of U* is substantially mitigated for the qubit models: for any SU(2) unitary U, U* = sigma_y U sigma_y, so the conjugate encoding can be implemented with the same single use of U plus fixed Pauli gates, making the MCM resource-fair for qubit systems. However, the experimental section reports no raw detection counts, no error bars on the plotted WMSE points, no state-preparation fidelity calibration, and no analysis code. With only 500 qubits per trial, the statistical uncertainty on each WMSE estimate is consequential, and without those data one cannot judge whether the claimed agreement with the QCRB lines is meaningful. Because the paper's central claim is an experimental demonstration, this lack of verifiable evidence is the most load-bearing weakness.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper theoretically proposes and experimentally demonstrates two photonic quantum statistical models with global antiunitary symmetry (GAS): the mutually conjugate model (MCM), consisting of |ψλ⟩⊗|ψλ*⟩, and the ancilla-assisted mutually conjugate model (AAMCM). For estimating (θ,ϕ) of a qubit with weight matrix W=diag(1,sin²θ), the authors show that these models saturate the multi-parameter scalar quantum Cramér-Rao bound (QCRB) with explicit parameter-independent POVMs, while the MCM achieves half the weighted mean square error of the parallel-copy strategy. A single-photon experiment using polarization and path degrees of freedom reports WMSE values for the MCM, parallel model, AAMCM, and maximally entangled model, claiming agreement with the theoretical QCRB/HCRB curves.","tokens_in":9644,"tokens_out":9039,"duration_ms":89151,"significance":"The theoretical core is sound and valuable: the closed-form derivations of the Fisher information matrices and the explicit Bell-type POVMs provide a clear, parameter-free demonstration of how GAS resolves the incompatibility of multi-parameter estimation. The factor-of-two improvement of the MCM over the parallel model is a clean and falsifiable prediction, and for qubit systems the conjugate unitary U* can be implemented with a single use of U plus fixed gates, which mitigates the resource-cost concern that naturally arises for general U. However, the experimental evidence is not reported to the standard needed to verify the central experimental claim: there are no error bars, raw counts, fidelity calibrations, or statistical analysis. The significance of the paper therefore rests on whether the experimental data can be provided and quantified; as written, the experimental demonstration is unverified.","major_comments":[{"comment":"The manuscript reports no error bars, no raw detection counts, no number of repeated trials, no state-preparation fidelity calibration, and no analysis code for any of the four models. With m=500 qubits per trial, the statistical uncertainty of each WMSE point is substantial, and the plotted points alone cannot establish that the experiment \"matches\" the theoretical curves. Please provide uncertainty quantification (e.g., bootstrap over repeated experimental runs or Bayesian credible intervals), raw data for a representative set of parameters, and a precise description of how the Bayesian estimator was implemented (including the update rule and prior). This is load-bearing because the central claim is an experimental demonstration of QCRB saturation.","section":"Experimental setup and results, Figs. 3 and 4"},{"comment":"The authors restrict θ and ϕ to the interval [0, π/2] \"to ensure a one-to-one correspondence between estimators and measurement results.\" However, the theoretical saturation claim in Eq. (6) and the claimed optimality of the POVMs are stated in the text for the full parameter range θ∈[0,π], ϕ∈[0,2π]. Moreover, the QFIMs of both the MCM and the AAMCM are singular at θ=0 (and the MCM also at θ=π), so the notion of saturating the QCRB at those points is not well defined. The manuscript should state the exact parameter domain on which the experimental claim holds and clarify whether the one-to-one restriction reflects a limitation of the specific estimator, the POVM, or the model itself.","section":"Experimental setup and results, parameter restriction"},{"comment":"The experimental WMSE points are compared directly to asymptotic QCRB and HCRB lines computed from Fisher information. With m=500 and a Bayesian estimator using a uniform prior on a restricted interval, the expected mean square error contains finite-sample, prior, and boundary corrections that are not quantified. A valid comparison requires either deriving the finite-sample expected WMSE of the implemented estimator (analytically or by simulation) or demonstrating numerically that all finite-size corrections are below the resolution of the plot. As written, the statement that the experiment \"match[es] the theoretical predictions\" is not yet a falsifiable claim.","section":"Experimental setup and results, Fig. 3 comparison to asymptotic bounds"}],"minor_comments":[{"comment":"The reference list is disordered: Ref. [64] is inserted between [37] and [39], and the same reference appears twice (as [64] and again at the end). Please renumber and deduplicate the bibliography.","section":"References"},{"comment":"The abstract states the precision \"is improved at least twice compared to conventional encoding strategies,\" but the paper establishes exactly a factor of two relative to the parallel model, not a lower bound against all conventional strategies. Please either prove a general lower bound or qualify the statement to name the specific comparison.","section":"Abstract and Eq. (6)"},{"comment":"The legend contains \"MUB Theory\" but the text never defines what this line represents; please specify whether it is the asymptotic HCRB of the parallel model under the MUB-like POVM and how it was computed.","section":"Fig. 3 legend"},{"comment":"The axis labels and legend in Fig. 4 are garbled (\"2 2 QCRB of ...\"), making it impossible to read which curve corresponds to which model. Please regenerate the figure with clear labels.","section":"Fig. 4 labels"},{"comment":"For qubit systems, the conjugation property U* = σ_y U σ_y means the MCM encoding can be implemented with a single use of U plus fixed Pauli gates; the paper should state this explicitly in the main text, since the discussion of the no-go theorem for general U may otherwise leave readers with the impression that the MCM advantage is not resource-fair even for the qubit platform demonstrated here.","section":"Summary and discussions"}],"recommendation":"major_revision","confidential_remarks":"The analytical part of this paper is solid and would be a useful contribution on its own. The experimental section, however, is currently not verifiable because it lacks uncertainty quantification and raw data. If the authors can provide the missing data and a proper statistical comparison, the paper would likely meet the standard for a strong letter. I would also check whether the journal's data-availability policy requires the raw counts to be deposited; if so, this should be a hard requirement for acceptance. The reference-numbering issue and missing figure labels are easy to fix but should be corrected in revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a solid theory result with an under-documented experiment. The analytical core holds up. For the MCM state |ψ>|ψ*>, the Bell POVM gives a classical Fisher information matrix equal to twice the QFIM, so the multiparameter QCRB is saturated without trade-off. The AAMCM QFIM and its separable POVM also check out, and the factor-of-two comparison with the parallel HCRB follows from Eq. (6). The weak-commutativity-to-GAS theorem is imported from Miyazaki and Matsumoto [37], so the conceptual novelty is modest, but the two concrete models, their optimal POVMs, and the photonic implementation are new. No fitted parameters enter the derivation; the benchmarks are independent analytical models. The experiment is well designed: single photons, polarization and path degrees of freedom, and deterministic collective measurements via quantum walks. The plotted mean-square errors are consistent with the theoretical curves. That is all one can say. There are no raw counts, no error bars on the weighted MSE points, no state-preparation fidelity numbers, and no analysis code. With 500 qubits per trial, the statistical uncertainty on each point is non-negligible, so the agreement with the QCRB lines is not independently assessable. The parameter range is restricted to θ,ϕ ∈ [0,π/2] to avoid estimator ambiguity from periodic probabilities; that is disclosed and reasonable, but it means the 'no trade-off' claim is demonstrated only on a subset. The Bell POVM also has singular Fisher information at some parameter values, and those non-regular points are not analyzed. The cost of preparing the conjugate state is a real concern for general systems, and the paper honestly flags the no-go theorem for U^{-1} in the Summary. For the qubit models actually demonstrated, U* = σ_y U σ_y, so conjugate encoding costs the same as U plus two fixed Pauli gates; the resource-fairness worry is therefore minor here. The citation pattern is appropriate, giving clear credit to the antecedent theory. Who this is for: people working on multiparameter quantum estimation, especially the QCRB-versus-HCRB gap, and experimental quantum optics groups. It deserves a serious referee. The referee should request raw data, error bars or credible uncertainty quantification, and ideally the analysis code before accepting the experimental claim. My verdict: the theory is likely correct, the experiment is likely fine, but the evidence as presented is not yet complete.","headline":"The analytical core is sound and the experiment is plausible, but without raw data and error bars the experimental claim is not yet independently verifiable; worth a serious referee.","tokens_in":10148,"tokens_out":3737,"would_cite":true,"duration_ms":34859,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P45","81P50","81P15"],"pacs":["03.65.Ta","03.65.Wj","42.50.Ex"],"model":"deepseek-v4-flash","headline":"Two-copy conjugate states can saturate the multi-parameter quantum Cramér-Rao bound for both parameters at once, with half the weighted mean squared error of two identical copies.","keywords":["multi-parameter quantum metrology","quantum Cramér-Rao bound","antiunitary symmetry","weak commutativity condition","mutually conjugate model","ancilla-assisted model","photonic experiment","quantum walk"],"falsifier":"Insert a controlled random relative phase between the two paths before the collective Bell measurement: the model then loses global antiunitary symmetry, and the paper's claim implies that no measurement can reach the scalar quantum Cramér-Rao bound value $\\mathrm{Tr}(WQ^{-1})/m$ for the MCM, while the unbroken experiment does; observing the predicted degradation would confirm the symmetry as the load-bearing resource.","tokens_in":9207,"feed_emoji":"🔬","tokens_out":10945,"duration_ms":97459,"temperature":0.7,"pith_summary":"The paper shows that requiring a family of quantum states to be invariant under an antiunitary operation removes the usual incompatibility between estimating different parameters at the same time. For the qubit state $|\\psi_\\lambda\\rangle = \\cos(\\theta/2)|0\\rangle + e^{i\\phi}\\sin(\\theta/2)|1\\rangle$, the two-copy mutually conjugate model $|\\psi_\\lambda\\rangle \\otimes |\\psi^*_\\lambda\\rangle$ obeys the weak commutativity condition everywhere, so the multi-parameter quantum Cramér-Rao bound is saturable for $\\theta$ and $\\phi$ simultaneously. The paper derives the chain $C^Q_\\lambda(|\\psi\\rangle\\otimes|\\psi^*\\rangle, W) = \\tfrac{1}{2} C^Q_\\lambda(|\\psi\\rangle,W) = \\tfrac{1}{4} C^H_\\lambda(|\\psi\\rangle,W) = \\tfrac{1}{2} C^H_\\lambda(|\\psi\\rangle^{\\otimes 2},W)$ with $W=\\mathrm{diag}(1,\\sin^2\\theta)$, and verifies it experimentally on single photons: the weighted mean squared error of the mutually conjugate model reaches the scalar QCRB and is half that of two identical copies. A second model with an ancilla, the AAMCM, reaches the same precision as the maximally entangled model while using only separable measurements.","feed_headline":"A state plus its conjugate doubles multi-parameter precision","feed_subtitle":"Photonic experiment reaches the ultimate precision bound for two angles at once, with no trade-off.","key_machinery":"The load-bearing object is the conjugation operator $\\vartheta$ and the global antiunitary symmetry condition $\\Theta\\rho_\\lambda\\Theta^\\dagger=\\rho_\\lambda$ for every $\\lambda$. A model with this symmetry has vanishing mean Uhlmann curvature, $U_{ij}=0$, which is precisely the weak commutativity condition needed to saturate the multi-parameter quantum Cramér-Rao bound. The symmetry also yields explicit optimal measurements: when the parameterized states decompose into parameter-independent eigenvectors with real coefficients, those eigenvectors form the optimal POVM; for the MCM this is a four-outcome Bell measurement, and for the AAMCM a separable measurement. The quantum-walk implementation realizes these POVMs deterministically in the polarization and path degrees of freedom of single photons.","core_discovery":"The central claim is that global antiunitary symmetry is a sufficient and practically available resource for simultaneous optimal estimation of multiple parameters. Concretely, the mutually conjugate model $\\{|\\psi_\\lambda\\rangle\\otimes|\\psi^*_\\lambda\\rangle\\}$ for estimating $(\\theta,\\phi)$ in a qubit satisfies the weak commutativity condition at every point of the parameter space, so an optimal measurement exists that attains the quantum Cramér-Rao bound for both parameters without trade-off. The paper also proves the quantitative improvement over the conventional parallel strategy: the scalar bound for the conjugate pair is exactly half the Holevo bound for two identical copies, a factor-of-two saving in weighted mean squared error. The photonic experiment reproduces the predicted weighted mean squared errors for the MCM and shows that the ancilla-assisted conjugate model matches the maximally entangled model's precision with a separable measurement.","pith_inferences":["Editorial inference: the factor-of-two statement is a resource-counting claim; counted in total channel uses rather than state copies, the advantage hinges on the price of implementing $U^*$, and the paper's own discussion shows that price is not always zero.","Editorial inference: the same symmetry argument should extend to tensor powers of conjugate pairs and to mixed states obeying $\\Theta\\rho_\\lambda\\Theta^\\dagger=\\rho_\\lambda$, potentially yielding parameter-independent optimal measurements in other estimation problems wherever a conjugation map exists.","Editorial inference: because the AAMCM needs only separable measurements, it is a plausible template for distributed or networked quantum sensing where collective measurements across nodes are costly; an adaptive multi-round version of the experiment would test that directly.","Editorial inference: the paper's link to 'nonlocality without entanglement' suggests the metrological gain is a geometric property of classically correlated quantum states, so analogous gains might appear in classical statistical models whose likelihood functions share the same symmetry."],"forward_implications":["The mutually conjugate strategy saturates the multi-parameter quantum Cramér-Rao bound for $(\\theta,\\phi)$ with no trade-off, a limit that two identical copies cannot reach with separable measurements.","The weighted mean squared error of the MCM is exactly half that of the parallel two-copy strategy, so the resource saving is a factor of two whenever preparing the conjugate state costs the same as preparing the original.","The ancilla-assisted mutually conjugate model reaches the same precision as the maximally entangled model while requiring only a separable measurement, replacing entanglement with antiunitary symmetry.","Any quantum statistical model with global antiunitary symmetry inherits compatible optimal measurements, giving a general recipe for parameter-independent optimal POVMs.","For real Hamiltonians, conjugation reduces to inverse evolution; a no-go theorem rules out exact inversion with a single use of $U$, though qubit systems admit exact deterministic time-reversal in special constructions, so the improvement is tied to how the conjugate operation is supplied."],"supporting_citations":[{"why":"Proves that parameterized states with global antiunitary symmetry obey the weak commutativity condition and gives explicit optimal POVM construction.","marker":"[37]"},{"why":"Gives the weak commutativity condition as the necessary and sufficient condition for saturating the multi-parameter quantum Cramér-Rao bound.","marker":"[24]"},{"why":"Supplies the Holevo Cramér-Rao bound relations, including the copy-number scaling used to compare the MCM with the parallel model.","marker":"[22]"},{"why":"Introduces the maximal entanglement model used as the precision benchmark that the AAMCM matches.","marker":"[9]"},{"why":"Provides the quantum-walk method used in the experiment to deterministically implement the collective optimal measurements.","marker":"[12]"},{"why":"States the no-go theorem that exact deterministic inversion of an unknown unitary with a single use is impossible, which the paper cites as the obstacle to implementing the conjugate model for real Hamiltonians.","marker":"[52]"}],"fun_headline_variants":["Antiunitary symmetry achieves multi-parameter quantum limit","Conjugate states reach Cramér-Rao bound for two parameters","Symmetry doubles multi-parameter precision without trade-off","Antiunitary symmetry doubles multi-parameter precision"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claimed factor-of-two advantage assumes that producing the conjugate state $|\\psi^*_\\lambda\\rangle$ costs no more than producing $|\\psi_\\lambda\\rangle$; if conjugation is expensive or unavailable, the comparison against the parallel-copy strategy is not resource-fair.","fun_headline_variants_meta":{"raw":{"variants":["Antiunitary symmetry achieves multi-parameter quantum limit","Conjugate states reach Cramér-Rao bound for two parameters","Symmetry doubles multi-parameter precision without trade-off","Antiunitary symmetry doubles multi-parameter precision"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00094,"raw_usage":{"total_tokens":3962,"prompt_tokens":834,"completion_tokens":3128,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":450,"completion_tokens_details":{"reasoning_tokens":3063}},"tokens_in":450,"tokens_out":3128,"duration_ms":23171,"temperature":1.0,"reasoning_tokens":3063,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:44:25.620144+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Insert a controlled random relative phase between the two paths before the collective Bell measurement: the model then loses global antiunitary symmetry, and the paper's claim implies that no measurement can reach the scalar quantum Cramér-Rao bound value $\\mathrm{Tr}(WQ^{-1})/m$ for the MCM, while the unbroken experiment does; observing the predicted degradation would confirm the symmetry as the load-bearing resource.","supporting_citations":[{"cited_title":"Chaichian and R","cited_arxiv_id":null,"evidence_quote":"Proves that parameterized states with global antiunitary symmetry obey the weak commutativity condition and gives explicit optimal POVM construction."},{"cited_title":"Hou, J.-F","cited_arxiv_id":null,"evidence_quote":"Introduces the maximal entanglement model used as the precision benchmark that the AAMCM matches."},{"cited_title":"Liu and H","cited_arxiv_id":null,"evidence_quote":"Provides the quantum-walk method used in the experiment to deterministically implement the collective optimal measurements."},{"cited_title":"Chiribella and D","cited_arxiv_id":null,"evidence_quote":"States the no-go theorem that exact deterministic inversion of an unknown unitary with a single use is impossible, which the paper cites as the obstacle to implementing the conjugate model for real Hamiltonians."}],"review_version":1}