{"id":"70952971-4a9e-47a8-b8d2-8e23eb89ef28","arxiv_id":"1909.01367","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For pure bipartite qutrit states, the paper derives and experimentally applies analytic relations between statistical correlation measures and the entanglement measures negativity and entanglement of formation.","lead":"This paper reports a way to measure how entangled two three-level quantum systems (qutrits) are, using only a few correlation measurements instead of full state reconstruction. The authors apply it to pairs of photons generated from a patterned crystal and find their state is close to, but not exactly, maximally entangled.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Measured σx correlation matrix violates the pure-state symmetry required by Eq. (8); the reported N and EOF values are not established.","rationale":"The reader's weakest assumption identifies the purity of the experimental state as the key vulnerability. My analysis sharpens this concern: the paper's own data contradict the pure-state assumption. For a pure bipartite qutrit, the generalized σx-basis joint probabilities have a specific symmetry (dependence only on j+k mod 3). The representative matrix in Table I clearly violates this symmetry, with deviations up to 0.084 between entries that should be equal. This is not a matter of missing a purity estimate; it is a direct inconsistency between the assumed state family and the measured correlation matrix. Furthermore, the reported PCC of 0.848(2) cannot be reproduced from the printed matrix under the eigenvalue assignment explicitly stated in the text; the assignment (0, -1, 1) is required instead. These two issues together mean the experimental quantification of entanglement, the central claim of the paper, is not internally consistent. The theoretical relations for pure states appear correct, and the experimental method may be salvageable with full state tomography or direct purity estimation, but as presented the central claim fails. I therefore recommend rejection of the current form, with the possibility of reconsideration after the authors provide a direct characterization of the state (e.g., two-qutrit tomography) and clarify the eigenvalue labeling. This is a more severe verdict than the reader's conditional acceptance because the listed problems are not merely omissions; they are contradictions within the reported data.","tokens_in":14419,"tokens_out":13960,"duration_ms":114996,"concrete_test":"Reanalyze the five measured σx correlation matrices: test the pure-state equalities P11 = P23 = P32, P12 = P21 = P33, and P13 = P31 = P22 with propagated Poisson errors from the raw coincidence counts. If any equality is violated by more than 3σ in a consistent direction, the pure-state assumption underlying Eq. (8) fails. In the same reanalysis, recompute the PCC from the printed matrix using the eigenvalue assignment stated in the text (0, 1, -1 for x1, x2, x3) and compare with the reported 0.848(2); a mismatch would confirm that the eigenvalue ordering or the matrix is misreported, invalidating the extracted negativity.","verdict_should_be":"REJECT","load_bearing_attack":"The central experimental claim relies on Eq. (8), |C_A1B1| = N, which is derived for pure bipartite qutrit states in Schmidt form. For such a pure state, the joint probabilities in the generalized σx basis, P(b_j,b_k) = (1/9)|Σ_i c_i ω^{-(j+k)i}|², depend only on j+k mod 3. Hence the 3×3 probability matrix must satisfy the equalities P11 = P23 = P32, P12 = P21 = P33, and P13 = P31 = P22. The representative σx matrix in Table I (right) violates these equalities: for example, P11 = 0.344, P23 = 0.260, and P32 = 0.302, a spread of 0.084 that far exceeds Poisson counting uncertainties. This is direct evidence that the prepared state is not a pure state of the assumed form, so Eq. (8) cannot be used to identify the measured PCC with the negativity. Additionally, using the eigenvalue assignment stated in the text (0, 1, and -1 for x1, x2, x3), the printed matrix yields |C| ≈ 0.54, not the reported 0.848(2); the reported value is reproduced only by the assignment (0, -1, 1). This internal inconsistency indicates a misreporting of either the eigenvalue ordering or the matrix itself. The 94% visibility and the agreement between N from PCC and MP do not resolve the issue, since both derivations assume pure states and the two measures are extracted from overlapping data. Consequently, the experimental values N = 0.848 ± 0.027 and EOF = 1.23 ± 0.01, and the claimed non-equivalence of entanglement measures, are not supported by the data as presented.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript claims three advances: analytical relations between statistical correlation measures and entanglement measures for pure bipartite qutrits, their experimental use to quantify entanglement in spatially correlated photonic qutrits, and the first experimental demonstration of non-equivalence between entanglement measures in higher dimensions. The central theoretical relations are Eq. (8), |C_A1B1| = N, where the Pearson correlation coefficient in a generalized sigma-x basis equals negativity; Eq. (11), relating mutual predictability to negativity; and Eqs. (14)-(15), identifying mutual information in the computational basis with entanglement of formation for pure states. The experiment uses a triple-slit modulated SPDC source, measures correlation matrices in image and focal planes, and reports N = 0.848 +/- 0.027 from PCC, N = 0.849 +/- 0.020 from MP, and EOF = 1.23 +/- 0.01. From these values the paper infers ~15% and ~24% deviations from the maximally entangled state and claims a non-equivalence of entanglement measures.","tokens_in":14736,"tokens_out":18651,"duration_ms":168597,"significance":"If the theoretical and experimental claims were fully established, the paper would be a useful contribution: it gives compact, analytic pure-state relations connecting directly measurable correlation statistics to negativity and entanglement of formation, and it applies them to a scalable spatial-bin qutrit source. The theoretical derivations for pure states are sound, and the Appendix material on monotonicity and non-equivalence is a strength. The paper is also careful to present representative correlation matrices and to report uncertainties from repeated measurements. However, the experimental conclusions rest on a pure-state assumption that is not directly verified and appears to be contradicted by the measured matrices; the EOF extraction is particularly problematic. The significance of the experimental demonstration is therefore conditional on a reanalysis or additional state characterization.","major_comments":[{"comment":"The measured focal-plane correlation matrix does not satisfy the symmetry required by Eq. (8) for a pure Schmidt state. For such a state, the generalized sigma-x joint probabilities depend only on (j+k) mod 3, so the equalities P11 = P23 = P32, P12 = P21 = P33, and P13 = P31 = P22 must hold. In the representative matrix these are violated by much more than Poisson counting uncertainties: for example, P11 = 0.344, P23 = 0.260, and P32 = 0.302, a spread of 0.084; likewise P12 = 0.017, P21 = 0.008, and P33 = 0.017. This is direct evidence that either the state is mixed or the effective measurement basis deviates from the ideal generalized sigma-x basis. In either case, identifying the measured PCC with the negativity through Eq. (8) is not justified as presented. The authors should provide a quantitative purity estimate, a calibration of the implemented measurement operators, or an error analysis showing that the symmetry violations are within the expected tolerance.","section":"Experimental scheme, Table I (right)"},{"comment":"The reported EOF value of 1.23 is not supported by the measured data as presented. Equation (14) equates mutual information with EOF only when the joint probabilities in the computational basis are exactly diagonal, p(ab) = c_i^2 delta_{ab}. The image-plane matrix in Table I (left) is not diagonal: off-diagonal entries such as 0.024, 0.014, 0.006, and 0.002 are nonzero and comparable to a few percent of the diagonal entries. Computing the classical mutual information from the full noisy joint distribution gives a number that is not an entanglement measure. Indeed, evaluating EOF from the diagonal entries alone gives approximately 1.56, whereas the reported 1.23 is close to the mutual information of the full image-plane matrix. The text is also ambiguous about whether the MI was computed from the image-plane or focal-plane data, because x1, x2, x3 and y1, y2, y3 were previously defined for the sigma-x basis. The authors need to specify the exact input matrix, justify why the non-diagonal terms can be ignored or corrected, and either provide a valid EOF estimate or withdraw the EOF-based non-equivalence claim.","section":"Relating Mutual Information with Entanglement of Formation; Experimental scheme"},{"comment":"The pure-state assumption is not adequately supported. The consistency between N from PCC and N from MP is presented as a purity check, but both quantities are extracted from the same focal-plane correlation matrix (the MP uses three entries of the same matrix that enters the PCC computation), so the agreement does not independently certify purity. The 94% visibility of the coincidence interference is a coherence indicator but not a direct estimate of the qutrit-state purity; it does not rule out the mixedness or measurement-basis errors implied by the symmetry violations in Table I. A direct purity estimate, such as a partial or full quantum state tomography or a fidelity bound, is needed before Eqs. (8) and (14) can be applied to the experimental data.","section":"Experimental scheme; Consistency check of purity"}],"minor_comments":[{"comment":"The derivation of Eq. (8) does not specify the numerical eigenvalues of the observables; later the text says eigenvalues 0, 1, and -1 are assigned. Please state explicitly that Eq. (8) holds for this particular eigenvalue assignment, since a different assignment (e.g., a different affine choice) would change the relation between PCC and negativity.","section":"Eq. (8) and eigenvalue assignment"},{"comment":"Appendix A derives the mutual predictability for sigma-x on both sides as 1/3, while the main text uses sigma-x on one side and its complex conjugate on the other to obtain Eq. (11). This distinction is easy to miss and should be highlighted in the main text to avoid confusion, especially because the experimental MP sum uses the off-diagonal x2-y3 and x3-y2 elements.","section":"Mutual Predictability definitions"},{"comment":"The correlation matrices in Table I are given without element-wise uncertainties or the normalization convention (e.g., whether rows, columns, or the total are normalized to unity). Providing these details would help the reader assess whether the symmetry violations in the focal-plane matrix are statistically significant.","section":"Table I"},{"comment":"The comparison of percentage deviations, Q_E and Q_N, depends on the different scales and normalizations of the two measures (log2(3) versus 1). The authors should discuss whether the quoted non-equivalence is a statement about the chosen normalization or about the more substantive non-monotonicity illustrated in Appendix C.","section":"Non-equivalence discussion"},{"comment":"There are typographical issues in the derivative formulas, such as the notation 'w.r.tc0' and the placement of logarithms, which should be cleaned up for readability.","section":"Appendix C, Eq. (43)"}],"recommendation":"major_revision","confidential_remarks":"The theoretical part of the paper is largely sound and the experimental source is interesting, but the quantitative experimental claims--especially the EOF value and the non-equivalence conclusion--are not supported by the data as presented. The symmetry violation in the focal-plane matrix and the non-diagonal image-plane matrix are concrete, checkable problems. I would like to see either a revised manuscript that addresses these with additional characterization (purity estimate, measurement-basis calibration, corrected EOF extraction) or a substantially weakened set of claims. This is not a reject because the analytic relations for pure states are correct and the experimental framework is potentially salvageable with a reanalysis."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the analytical relation |C_A1B1| = N for pure bipartite qutrits (Eq. 8) looks correct to me, and the use of Pearson correlation for this purpose is genuinely new. Second, the experimental numbers do not support the main claim: the measured sigma_x correlation matrix in Table I (right) is not consistent with a pure Schmidt state, so equating the measured PCC to the negativity is not justified.\n\nThe experimental setup is clever, and the idea of quantifying entanglement from a handful of correlation measurements is practically appealing. The data were collected carefully, and the paper is clearly written. But the central quantitative results have a load-bearing problem. For a pure Schmidt state, the joint probabilities in the generalized sigma_x basis depend only on j+k mod 3. The right-hand matrix of Table I should therefore satisfy P(y1,x1) = P(y2,x3) = P(y3,x2). The measured values are 0.344, 0.260, and 0.302, a spread far larger than the Poisson counting errors. The prepared state is not the pure state the derivation assumes. The agreement between N from PCC and N from MP is not an independent consistency check, because both are computed from the same matrix.\n\nThe mutual-information route to EOF is also under-specified. Equation (14) is derived for measurements in the computational basis, but the text suggests the sigma_x positions were used. As written, I cannot reproduce the quoted EOF = 1.23. The stated 94% interference visibility does not fix this; visibility is a rough proxy for coherence, not a quantifiable bound on mixedness that would validate Eq. (8).\n\nThe claim of first experimental demonstration of non-equivalence between entanglement measures is weaker than presented. N and EOF are different functions of the state, and it is unsurprising that percentage deviations from the maximally entangled state differ. One small point: the stress-test worry about eigenvalue ordering does not hold. Flipping the signs of all eigenvalues leaves |C| unchanged, so the printed matrix is consistent with the stated assignment.\n\nWho is this for? Researchers working on measurement-efficient entanglement quantification in higher-dimensional systems will find the theory section useful, and the experimental platform is interesting. But as it stands, the reported values of N and EOF are not established. I would send this to peer review with the expectation of major revision: authors should provide a direct purity estimate or full state characterization, correctly derive MI from the appropriate basis, and temper the quantitative claims. I would not cite the numbers in their current form.","headline":"The theory relation is interesting, but the experimental values don't hold because the measured matrix violates the pure-state symmetry the method relies on.","tokens_in":15270,"tokens_out":12100,"would_cite":false,"duration_ms":96315,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P40","81P68"],"pacs":["03.67.Mn"],"model":"deepseek-v4-flash","headline":"For pure bipartite qutrits, the Pearson correlation coefficient in the generalized $\\sigma_x$ basis equals the negativity; the paper uses this identity to quantify a spatial-bin photon pair at $N=0.85$ and $EOF=1.23$.","keywords":["entanglement quantification","qutrit","Pearson correlation coefficient","negativity","entanglement of formation","mutual information","spatial-bin photonic qutrits","spontaneous parametric down-conversion"],"falsifier":"Perform a full tomographic reconstruction of the two-qutrit state at the same source settings, compute its negativity and entanglement of formation directly from the density matrix, and compare with $0.848\\pm0.027$ and $1.233\\pm0.012$; disagreement beyond statistical error would show the pure-state formulas are not valid for this state.","tokens_in":14207,"feed_emoji":"🔗","tokens_out":10646,"duration_ms":100290,"temperature":0.7,"pith_summary":"Higher-dimensional entanglement has mostly been certified through bounds, not measured. This paper closes that gap for pure bipartite qutrits by deriving exact analytical links between three statistical correlation measures and two standard entanglement measures, and by using those links to quantify the entanglement in a photonic spatial-bin qutrit source. The central identity is $|C_{A_1B_1}|=N$: the Pearson correlation coefficient of outcomes in the generalized $\\sigma_x$ basis equals the negativity of a Schmidt-decomposed two-qutrit state. A mutual-predictability measurement in a conjugate basis gives the same negativity, and computational-basis mutual information equals the entanglement of formation. Applying these to the generated photon pairs yields $N=0.85\\pm0.03$ and $EOF=1.23\\pm0.01$, and the two measures disagree about the distance to the maximally entangled state: about 15% versus 24%.","feed_headline":"One correlation measurement equals qutrit negativity exactly","feed_subtitle":"Applied to photon pairs, it gives negativity 0.85 and entanglement of formation 1.23.","key_machinery":"The load-bearing object is the mutually unbiased pair of qutrit bases used for the measurements. The generalized $\\sigma_x$ basis $\\{|b_0\\rangle,|b_1\\rangle,|b_2\\rangle\\}$ is built from discrete Fourier combinations of the computational basis, $|b_j\\rangle=(|0\\rangle+\\omega^j|1\\rangle+\\omega^{2j}|2\\rangle)/\\sqrt{3}$ with $\\omega=e^{2\\pi i/3}$, while the computational basis serves as the $\\sigma_z$-like basis. When Alice and Bob both measure the observable whose eigenstates are the $|b_j\\rangle$'s, the Pearson correlation coefficient collapses to the negativity $N$, and this identity is what converts a few coincidence counts into a quantitative entanglement value. The companion formulas for mutual predictability and mutual information extend the same data to a second, independent entanglement measure.","core_discovery":"The paper's claim is that entanglement in pure bipartite qutrit states is not only certifiable but directly readable from a few correlation measurements. For any state written in Schmidt form, the Pearson correlation coefficient for the generalized $\\sigma_x$ observable on both sides is exactly the negativity: $|C_{A_1B_1}|=N$, with $N=c_0c_1+c_0c_2+c_1c_2$. Measuring $\\sigma_x$ on one side and its complex conjugate on the other gives mutual predictability $C=\\frac{1}{3}(1+2N)$, and measuring in the computational basis gives mutual information $I_{AB}=-\\sum_i c_i^2\\log_2 c_i^2$, which is exactly the entanglement of formation. The experiment records these correlations for pump-beam-modulated spontaneous parametric down-conversion, certifies entanglement through focal-plane interference, and reports three consistent values ($N=0.848\\pm0.027$ from PCC, $N=0.849\\pm0.020$ from MP, $EOF=1.233\\pm0.012$), demonstrating the first experimental non-equivalence of the two measures across dimensions.","pith_inferences":["A direct tomographic reconstruction of the same source, compared with the correlation-derived numbers, would settle how much residual mixedness biases the pure-state formulas; the paper does not itself run that test.","The identity $|C_{A_1B_1}|=N$ suggests a generic recipe: any platform with access to a Fourier-type basis, whether time-bin, orbital angular momentum, or frequency, could use the same Pearson-correlation measurement as an entanglement meter for pure qudits.","The observed non-equivalence between negativity and EOF implies that resource-theoretic claims for high-dimensional protocols should be measure-specific; a state that looks closer to maximal under one measure may look farther under the other.","Tuning the slit parameters to vary the Schmidt coefficients would allow a direct experimental scan of the predicted non-monotonic relationship between $N$ and EOF, connecting the demonstration to the theoretical analysis in the appendix."],"forward_implications":["Pure qutrit entanglement can be quantified from a small set of coincidence measurements, avoiding the full tomographic reconstruction that scales poorly with dimension.","The generalized $\\sigma_x$ correlation measurement, the conjugate-basis predictability measurement, and the computational-basis mutual-information measurement each give a certified entanglement value, with the two negativity routes agreeing within error.","Negativity and entanglement of formation are not equivalent reporters of qutrit entanglement: they place the prepared state at roughly 15% and 24% deviation from maximal entanglement, so claims about the quality of a high-dimensional state should specify which measure is being used.","The mutual-information/EOF identity is derived for any pure qudit dimension, so the same correlation-measurement approach can be carried to higher-dimensional spatial-bin systems without a new theoretical derivation for EOF."],"supporting_citations":[{"why":"Proposes the Pearson-correlation-based entanglement criterion that this paper proves for pure qutrits and uses as its starting point.","marker":"[36]"},{"why":"Provides the generalized $\\sigma_x$-basis formalism and the broader mixed-state analysis of higher-dimensional entanglement that underpins the qutrit derivation.","marker":"[37]"},{"why":"Shows that focal-plane interference in coincidence measurements certifies entanglement, which the experiment uses as its certification step.","marker":"[47]"},{"why":"Introduces the pump-beam modulation technique that generates the spatial-bin photonic qutrit pairs measured here.","marker":"[52]"},{"why":"Supplies the measurement-operator description for slit-image and far-field detection used to implement generalized $\\sigma_z$ and $\\sigma_x$ bases.","marker":"[53]"},{"why":"Defines mutual predictability and gives the formula the paper uses to connect it to negativity.","marker":"[61]"},{"why":"Defines the negativity measure that appears in the central equalities.","marker":"[62]"},{"why":"Introduces the $Q_E$ and $Q_N$ deviation measures and the two-qubit non-equivalence analysis that this paper extends to qutrits.","marker":"[54]"}],"fun_headline_variants":["Exact qutrit entanglement from correlation statistics","Qutrit entanglement measured exactly via correlations","First exact quantification of qutrit entanglement","Correlations reveal non-equivalent qutrit entanglement measures"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire quantitative bridge assumes the prepared pair is a pure bipartite qutrit state; if the actual state is noticeably mixed, the measured correlation coefficients no longer equal negativity and entanglement of formation, and the reported numbers inherit an unquantified error.","fun_headline_variants_meta":{"raw":{"variants":["Exact qutrit entanglement from correlation statistics","Qutrit entanglement measured exactly via correlations","First exact quantification of qutrit entanglement","Correlations reveal non-equivalent qutrit entanglement measures"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000596,"raw_usage":{"total_tokens":2797,"prompt_tokens":961,"completion_tokens":1836,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":577,"completion_tokens_details":{"reasoning_tokens":1777}},"tokens_in":577,"tokens_out":1836,"duration_ms":14304,"temperature":1.0,"reasoning_tokens":1777,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:21:04.604637+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform a full tomographic reconstruction of the two-qutrit state at the same source settings, compute its negativity and entanglement of formation directly from the density matrix, and compare with $0.848\\pm0.027$ and $1.233\\pm0.012$; disagreement beyond statistical error would show the pure-state formulas are not valid for this state.","supporting_citations":[{"cited_title":"Jebarathinam, D","cited_arxiv_id":null,"evidence_quote":"Provides the generalized $\\sigma_x$-basis formalism and the broader mixed-state analysis of higher-dimensional entanglement that underpins the qutrit derivation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that focal-plane interference in coincidence measurements certifies entanglement, which the experiment uses as its certification step."},{"cited_title":"Kolenderski, U","cited_arxiv_id":null,"evidence_quote":"Supplies the measurement-operator description for slit-image and far-field detection used to implement generalized $\\sigma_z$ and $\\sigma_x$ bases."},{"cited_title":"Revisiting comparison between entanglement measures for two-qubit pure states","cited_arxiv_id":"1907.09268","evidence_quote":"Introduces the $Q_E$ and $Q_N$ deviation measures and the two-qubit non-equivalence analysis that this paper extends to qutrits."}],"review_version":1}