{"id":"abdd2ebe-40d0-4372-ac24-db28e76aa21f","arxiv_id":"1909.00755","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In qubit postselected measurements, anomalous values, enhanced Fisher information, and contextuality are not equivalent outside the strict weak-value regime.","lead":"Experiments on single photons show that three quantum signatures (anomalous values, enhanced Fisher information, and contextuality) in postselected measurements do not always appear together outside the very weak measurement limit. The paper gives a mathematical reason why enhanced Fisher information is not tied to anomalous values and shows anomalous values without contextuality.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fig. 3's Pusey-inequality test uses a non-optimal disturbance probability pd; with the minimal allowed pd the claimed decoupling of anomalous values and contextuality is not yet established.","rationale":"I read the central claim as a negative universal: for qubit postselected measurements outside the strict weak limit, anomalous values, enhanced Fisher information, and contextuality are not in one-to-one correspondence. The paper supports this by showing, among other things, that anomalous values can occur while the Pusey inequality (5) is satisfied (Fig. 3). The load-bearing point I found is not the scope of the measurement family (the reader's concern) but the validity of the Pusey bound used. The Appendix's decomposition has a smaller, still valid pd, so the bound in Eq. (5) is not the strongest consequence of non-contextuality. This weakens the experimental pillar of the no-mapping claim. The fix is straightforward: recompute with pd_min. I therefore keep the reader's CONDITIONAL verdict—the theoretical core may survive, but the current experimental evidence for the decoupling is not yet convincing. This is a different concern from the reader's, hence disagreement on the weakest assumption.","tokens_in":7906,"tokens_out":33108,"duration_ms":286115,"concrete_test":"Recompute I0 in Eq. (5) for the data of Fig. 3 using pd,min=(1−√(1−κ²))/2 instead of the value used in the paper, with the same experimental p0, pφ, and κ=0.335. If every point remains negative, the experimental conclusion is robust; if any point becomes non-negative, the demonstration of anomaly-without-contextuality fails. As a check, verify numerically that S−pd,min|φ⟩⟨φ| is positive semidefinite for the actual φ=± states.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Eq. (5) and Fig. 3 are meant to show that anomalous values can appear without implying contextuality. The proof in the Appendix requires a valid decomposition S=(1−pd)|φ⟩⟨φ|+pdE_d for S=M0|φ⟩⟨φ|M0†+M1|φ⟩⟨φ|M1†. For the measurement operators in Eq. (1), direct algebra gives that S−c|φ⟩⟨φ| is positive semidefinite for every c≤(1+√(1−κ²))/2, so the strongest (smallest) disturbance probability is pd,min=(1−√(1−κ²))/2. The paper instead uses pd=1−√(1−κ²) in Eq. (5), which is valid but weaker by (1−√(1−κ²))/(2pφ). The Fig. 3 caption says the data are evaluated under 'the best circumstances for disproving the non-contextual model', but a strictly stronger test is available. If any data point satisfies −(1−√(1−κ²))/(2pφ) < I0 < 0, it would violate the optimal Pusey inequality, so the observed absence of violation—and the claimed decoupling of anomalous values and contextuality—would not be established.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript considers postselected generalized measurements on qubits, implemented with a qubit meter coupled through a C-Sign gate, and studies the relations among three quantum signatures: anomalous values of the postselected observable, enhanced conditional Fisher information, and contextuality as probed by the Pusey inequality. The main theoretical results are Eq. (6), which expresses the postselected Fisher information in terms of kappa*sigma_w, and an extension of Pusey's non-contextuality inequality to the discrete measurement operators of Eq. (1). The paper reports an experiment measuring weak values, Fisher-information-related variances, and the Pusey quantity for two postselections, and concludes that outside the strict weak-value regime there is no one-to-one mapping among the three phenomena. The central claim is that for qubits these effects must be assessed independently.","tokens_in":8142,"tokens_out":10269,"duration_ms":84361,"significance":"If established, the result is a useful clarification of the relations among anomalous values, enhanced Fisher information, and contextuality in postselected quantum measurements. The derivation of Eq. (6) is clean and the extension of the Pusey inequality to a discrete qubit-meter setting is a valuable technical contribution. The experimental implementation is relevant and the paper explicitly acknowledges limitations about the scope of the no-one-to-one claim. However, the experimental support for the contextuality decoupling is weakened by the use of a non-optimal disturbance probability, and the Fisher-information comparison lacks uncertainty estimates. With these points addressed, the manuscript would be suitable for publication.","major_comments":[{"comment":"The Pusey-inequality test uses pd = 1 - sqrt(1 - kappa^2), but direct algebra on the operators of Eq. (1) shows that S - c|phi><phi| is positive semidefinite for every c <= (1 + sqrt(1 - kappa^2))/2, so the smallest valid disturbance probability is pd,min = (1 - sqrt(1 - kappa^2))/2. The inequality with pd,min is strictly stronger, and the caption of Fig. 3 claims the data are evaluated under 'the best circumstances for disproving the non-contextual model,' which is not the case with the reported pd. Concretely, I0_opt = I0_reported + (1 - sqrt(1 - kappa^2))/(2 p_phi), so any data point with -(1 - sqrt(1 - kappa^2))/(2 p_phi) < I0_reported < 0 would violate the optimal Pusey inequality. The conclusion that anomalous values can appear without contextuality is therefore not established unless the data are reanalyzed with the optimal pd or a clear justification for the weaker choice is provided.","section":"Eq. (5), Appendix, Fig. 3"},{"comment":"The abstract and introduction state that 'when performing generic postselected measurements there exist no one-to-one mapping' among anomalous values, enhanced Fisher information, and contextuality. The proof and the experimental demonstration, however, are confined to the specific two-outcome qubit-meter family of Eq. (1) with postselection after the measurement. The authors acknowledge some caveats later, but the abstract's use of 'generic' overstates the scope. Please qualify the claim in the abstract and conclusions so that it is clear the result is established for this family and this ordering, and that other measurement strategies may behave differently.","section":"Abstract and Conclusions"},{"comment":"Satisfying the Pusey inequality is a necessary condition for the existence of a non-contextual model, not a sufficient one. The statement in the Conclusions that 'a non-contextual model becomes appropriate' is therefore too strong: the experimental data can at most show that this particular inequality does not rule out non-contextual models. Rephrase to avoid implying that a non-contextual model has been constructed or established, and make clear that the conclusion is about the failure of one specific test.","section":"Fig. 3 and Conclusions"},{"comment":"The comparison between measured variances and the Cramer-Rao bound is central to the Fisher-information discussion, but Table I lists no uncertainties on the measured Delta^2 theta values, making it difficult to judge whether the differences from sigma_CR are significant. In addition, the theoretical curves in Fig. 2 rely on parameters v = 0.78, T_H = 0.98, and T_V = 0.34, whose origin is not stated in the text; the text also mentions a visibility limited to 97%, which is inconsistent with v = 0.78. Please provide error bars for Table I and specify how the imperfection parameters were determined.","section":"Table I and Fig. 2"}],"minor_comments":[{"comment":"In the Conclusions, 'more pondering' should read 'more pondered' or 'considered more carefully'.","section":"Conclusions"},{"comment":"The notation for pd is inconsistent between Eq. (5), where it is given as 1 - sqrt(1 - kappa^2), and the Appendix, where the printed text can be read as 1 - 2*sqrt(1 - kappa^2); please harmonize the notation and ensure the intended expression is unambiguous.","section":"Eq. (5) and Appendix"},{"comment":"The figure caption should state whether the error bars on I0 include propagation of both the coincidence-count statistics and the uncertainty on kappa.","section":"Fig. 3"},{"comment":"The acronyms PPBS and C-Sign should be defined at first use, and the relation between the meter angle mu and kappa = sin(4 mu) should be stated clearly in the main text rather than only in the figure caption.","section":"Experimental setup"}],"recommendation":"major_revision","confidential_remarks":"The suboptimal choice of pd in the Pusey test is a load-bearing issue: if the optimal inequality is violated by any of the measured data points, the claimed decoupling of anomalous values from contextuality would be contradicted. The analysis can likely be fixed within the scope of the manuscript by recomputing Fig. 3 with the optimal pd and adjusting the conclusions accordingly. I recommend asking the authors to perform this reanalysis and to report both versions of the Pusey quantity. The other major comments concern scope and presentation, which are addressable by revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core of this paper is a useful correction: for qubit postselected measurements, anomalous values, enhanced Fisher information, and contextuality only coincide in the strict weak limit. Eq. (6) makes the Fisher information point cleanly—Fps depends on σw through κσw, so an anomalous value is neither necessary nor sufficient for enhanced precision. The extension of Pusey's inequality to finite-disturbance measurements with a qubit meter is a genuine contribution, and the Appendix derivation is clear.\n\nThe soft spots are experimental. The Pusey-inequality test in Fig. 3 uses pd = 1−√(1−κ²), but the optimal decomposition allows pd,min = (1−√(1−κ²))/2. Using the weaker, larger disturbance probability makes it easier to satisfy the non-contextual bound. The caption's claim that this is the 'best circumstances' for disproving the non-contextual model is not correct. With the tighter bound, a data point in the range −(1−√(1−κ²))/(2pφ) < I0 < 0 would actually violate Pusey, so the observed absence of violation is not conclusive. The Appendix also contains a likely typo stating pd = 1−2√(1−κ²), which is impossible for κ < √3/2; the intended formula is presumably pd = 1−√(1−κ²), but that is still non-optimal.\n\nThe other experimental weaknesses are the missing error bars in Table I and the unstated origin of the imperfection parameters (v=0.78, TH=0.98, TV=0.34) used in Fig. 2. The text mentions a visibility limited to 97% but the model uses v=0.78; that discrepancy needs explanation. Are these calibrated or fitted?\n\nThe theoretical argument holds up and is valuable. The experimental demonstration, however, needs stronger data analysis before the decoupling claim is experimentally established. For quantum foundations and weak-value metrology, this is a useful contribution that deserves peer review, but the revision should fix the Pusey test and provide proper error analysis.","headline":"The theoretical point about decoupling is sound and worth taking seriously, but the experimental test of the Pusey inequality uses a suboptimal disturbance parameter and the error analysis is too thin to fully support the experimental claims.","tokens_in":8679,"tokens_out":8279,"would_cite":true,"duration_ms":139719,"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":"For qubits, the paper shows that anomalous values, enhanced Fisher information, and contextuality are not the same phenomenon in generic postselected measurements; only in the strict weak-value limit do they coincide.","keywords":["weak measurements","anomalous weak values","Fisher information","contextuality","Pusey inequality","postselection","qubit metrology","generalized measurements"],"falsifier":"The claim would be falsified by an exhaustive numerical scan of the family of Eq. (1) at some fixed $\\kappa>0$ showing that the three conditions $|\\sigma_w|>1$, $F_{\\rm ps}>Q$, and a Pusey violation hold together for every preparation and postselection, since the paper asserts that no such one-to-one correspondence exists; the present experiment already demonstrates the decoupling for one strength.","tokens_in":7714,"feed_emoji":"⚛️","tokens_out":9297,"duration_ms":81753,"temperature":0.7,"pith_summary":"Postselected weak measurements on qubits are known to produce three effects: values outside the observable's spectrum, Fisher information that exceeds the standard quantum limit, and probabilities that can resist non-contextual models. This paper asks whether the three are the same underlying phenomenon. It shows they are not: for the generalized qubit measurements defined by Eq. (1), there is no one-to-one mapping among anomalous values, enhanced Fisher information, and contextuality. The link exists only in the strict weak-value limit, where the disturbance vanishes. A photonic experiment with intermediate measurement strength confirms the decoupling: anomalous values appear while the Pusey inequality remains satisfied, and the Fisher-information enhancement tracks the functional shape of the postselected value, not its anomaly.","feed_headline":"Anomalous values don't imply contextuality outside the weak limit","feed_subtitle":"Photonic qubit data: anomalous weak values pass the Pusey test, so each signature must be checked on its own.","key_machinery":"The machinery is the one-parameter family of measurement operators $M_0$, $M_1$ of Eq. (1), realized by a controlled-sign gate coupling the qubit to a meter qubit, with postselection after the interaction. The argument runs through two identities: the anomalous-value definition $\\sigma_w = (p^c_0-p^c_1)/\\kappa$, and the Fisher information $F_{\\rm ps} = \\kappa^2(\\partial_\\theta\\sigma_w)^2/(1-\\kappa^2\\sigma_w^2)$, whose denominator shows that only $\\kappa\\sigma_w$ matters. The third ingredient is the Pusey inequality $I_0<0$, extended in the appendix from continuous meters to this positive-operator-valued measurement, which supplies the quantitative test for non-contextual models, where measurement outcomes are predetermined independently of the context. These pieces let the paper scan states and show that the conditions fail to align.","core_discovery":"Within the family of qubit measurements (1) followed by postselection on $\\langle\\varphi|$, the paper establishes that the three signatures are independent. The anomalous value $\\sigma_w=(p^c_0-p^c_1)/\\kappa$ can exceed the $[-1,1]$ spectrum of the measured observable while the Pusey-type inequality (5) is satisfied, so no non-contextual model is ruled out; the experiment observes exactly this. The conditional Fisher information $F_{\\rm ps}=\\kappa^2(\\partial_\\theta\\sigma_w)^2/(1-\\kappa^2\\sigma_w^2)$ depends only on the combination $\\kappa\\sigma_w$, so its enhancement is a response to the changed functional dependence of $\\sigma_w$ on $\\theta$, not to the presence of anomalies. The paper's conclusion is a negative universal statement: apart from the strict weak-value regime, anomalous values, improved Fisher information, and contextuality must be assessed separately, and claims that couple them are ordering-dependent, as seen in the difference from the postselection-before-measurement scheme.","pith_inferences":["Beyond the paper, the qubit-specific decoupling leaves open whether higher-dimensional meters or multi-outcome postselection could re-establish a one-to-one link at finite strength; the paper's proof technique does not obviously extend to those cases.","The functional form $F_{\\rm ps}=\\kappa^2(\\partial_\\theta\\sigma_w)^2/(1-\\kappa^2\\sigma_w^2)$ suggests that metrological optimization should target the slope of $\\sigma_w(\\theta)$ rather than the magnitude of the anomaly, a strategy testable by varying $\\kappa$ continuously.","The order dependence the paper notes between its scheme and the postselection-before-measurement scheme implies that contextuality claims for postselected measurements are operationally relational: any comparison must specify the ordering of measurement and postselection."],"forward_implications":["For generic postselected qubit measurements, an anomalous value is not a witness of contextuality: the experiment records anomalous $\\sigma_w$ while the Pusey inequality stays satisfied.","A conditional Fisher information above the standard quantum limit does not certify an advantage per prepared event; the postselection probability cancels the gain in total resource counting.","The connection between anomalies, Fisher information, and contextuality is restored only in the strict weak-value limit $\\kappa\\ll1$, so claims valid there cannot be extrapolated to stronger measurements.","The generalized measurement can be viewed as an effective non-Hermitian (PT-symmetric) measurement, with postselection corresponding to ignoring outcomes, matching behaviour reported in other settings."],"supporting_citations":[{"why":"Defines weak values, establishing the weak-value limit that the paper contrasts with generic postselected measurements.","marker":"[1]"},{"why":"Supplies the qubit-based weak measurement implementation and the conditional-probability definition of anomalous values used in Eq. (3).","marker":"[2]"},{"why":"Provides the Pusey inequality for non-contextual models that the paper extends and tests.","marker":"[28]"},{"why":"Proves that postselection cannot beat the standard quantum Fisher information when total repetitions are counted, grounding the resource bound.","marker":"[31]"},{"why":"Shows that conditional Fisher information can exceed the quantum limit and serves as the baseline for the paper's tenuous-anomaly discussion.","marker":"[32]"},{"why":"Relates postselection, Fisher information, and contextuality in the opposite ordering, the scheme the paper distinguishes from its own.","marker":"[42]"},{"why":"Introduces the operator family of Eq. (1) that defines the generalized measurements under test.","marker":"[43]"}],"fun_headline_variants":["Anomalous values don't imply contextuality in postselected qubits","Qubit postselection: anomalous values, Fisher info, contextuality decoupled","Anomalous values pass Pusey test: signatures independent","No link between anomalous values and contextuality in qubit postselection","Fisher info boost not tied to anomalous values in qubit data"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the one-parameter family of qubit measurements followed by postselection, studied here, fairly represents generic postselected measurements; if another family restored a one-to-one link, the general claim would fail.","fun_headline_variants_meta":{"raw":{"variants":["Anomalous values don't imply contextuality in postselected qubits","Qubit postselection: anomalous values, Fisher info, contextuality decoupled","Anomalous values pass Pusey test: signatures independent","No link between anomalous values and contextuality in qubit postselection","Fisher info boost not tied to anomalous values in qubit data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000832,"raw_usage":{"total_tokens":3594,"prompt_tokens":872,"completion_tokens":2722,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":488,"completion_tokens_details":{"reasoning_tokens":2630}},"tokens_in":488,"tokens_out":2722,"duration_ms":17462,"temperature":1.0,"reasoning_tokens":2630,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:37:23.988971+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The claim would be falsified by an exhaustive numerical scan of the family of Eq. (1) at some fixed $\\kappa>0$ showing that the three conditions $|\\sigma_w|>1$, $F_{\\rm ps}>Q$, and a Pusey violation hold together for every preparation and postselection, since the paper asserts that no such one-to-one correspondence exists; the present experiment already demonstrates the decoupling for one strength.","supporting_citations":[{"cited_title":"consolidated measurement","cited_arxiv_id":null,"evidence_quote":"Defines weak values, establishing the weak-value limit that the paper contrasts with generic postselected measurements."},{"cited_title":"Avella, F","cited_arxiv_id":null,"evidence_quote":"Provides the Pusey inequality for non-contextual models that the paper extends and tests."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proves that postselection cannot beat the standard quantum Fisher information when total repetitions are counted, grounding the resource bound."},{"cited_title":"Dziewior, L","cited_arxiv_id":null,"evidence_quote":"Relates postselection, Fisher information, and contextuality in the opposite ordering, the scheme the paper distinguishes from its own."},{"cited_title":"Quantum Advantage in Postselected Metrology","cited_arxiv_id":"1903.02563","evidence_quote":"Introduces the operator family of Eq. (1) that defines the generalized measurements under test."}],"review_version":1}