{"id":"43a0abe0-4b70-4e07-b66a-79cd458ddd71","arxiv_id":"2504.13827","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":11,"one_line_summary":"A dispersive fit to e+e−→3π data yields F3π = 33.1(1.7) GeV^-3, consistent with the chiral anomaly prediction at the 5% level.","lead":"This paper extracts the chiral anomaly constant F3π, which controls the electromagnetic interaction of three pions, from electron-positron collision data using a dispersive fit. The result matches the theoretical Wess-Zumino-Witten prediction to about 5% precision, offering a new consistency check for hadronic physics and for the muon's anomalous magnetic moment.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Conformal polynomial truncation may not converge; the 5% precision claim relies on an unverified high-energy extrapolation of the sum rule.","rationale":"The reader's weakest assumption is that missing contributions above 1.8 GeV or insufficient conformal-polynomial order could shift C_A. This is exactly the load-bearing point. Table 3 makes the concern concrete: the central value moves by 0.048 from p_conf = 3 to p_conf = 4, which is the full quoted truncation error, and the paper's own statement that larger p_conf become 'increasingly unstable' suggests non-convergence. The poor global fit quality (p-value ~10^-5) reinforces that the representation is incomplete. These facts mean the 5% precision claim is fragile, but the qualitative result—C_A consistent with the chiral prediction—remains plausible and is corroborated by independent COMPASS and Primakoff determinations. The paper is honest about the dominant truncation uncertainty and about the tensions, so conditional acceptance with a requirement to demonstrate convergence of the conformal expansion is appropriate. The verdict thus remains CONDITIONAL, and my stress-test does not change the reader's assessment, though it sharpens the condition that should be met before final acceptance.","tokens_in":17894,"tokens_out":12777,"duration_ms":117227,"concrete_test":"Run the global fits to SND, CMD-2, and BaBar with p_conf = 5, 6, and 7, and record C_A/C_WZW at each order (the paper already shows p_conf = 5 in Fig. 1 but does not report the values or errors). If the difference between consecutive orders does not fall below approximately 0.02 (i.e., the order-to-order change shrinks to well within the quoted truncation error of 0.048) by p_conf = 7, or if the sequence continues to drift monotonically, the 5% precision and the associated error budget are not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central value C_A/C_WZW = 1.028(15)(48)(17)[53] is obtained by imposing the sum rule (2.11) at every fit step, so the anomaly is extracted as the integral of the imaginary part of the subtraction function. The imaginary part above s_inel = 1 GeV^2 is generated entirely by a truncated conformal polynomial C_p(q^2), which the data constrain only up to about 1.8 GeV. The truncation error is estimated from the spread across p_conf = 2, 3, 4, but Table 3 shows the sequence C_A/C_WZW = 1.038(11), 1.028(13), 0.980(16) for δ = 0, with the single step p = 3 to p = 4 already moving by 0.048, equal to the quoted truncation error. The paper itself notes in Sec. 3 that for larger p_conf 'the fits become increasingly unstable', i.e., the expansion has not demonstrably converged by p_conf = 4. Since the global fit p-value is around 10^-5 (Table 1), the representation is incomplete, and the sum rule integral may be biased either by missing high-energy contributions or by an over-flexible polynomial that absorbs data imperfections. The claimed 5% precision therefore depends on an unverified convergence of the conformal expansion, not just on a well-defined systematic envelope.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The authors extract the chiral anomaly F3π from e+e−→3π cross-section data using the dispersive Khuri–Treiman representation of the γ*→3π amplitude developed in earlier work. The central idea is to leave the subtraction constant α_A (proportional to the anomaly constant C_A) free in the fit, while imposing the sum rule (2.11) at every fit step. Fitting SND, CMD-2, and BaBar data yields F3π = 33.1(1.7) GeV^{-3}, i.e. C_A/C_WZW_A = 1.028(15)(48)(17)[53], in agreement with the WZW prediction at the 5% level. The paper also fits the recent Belle II data and finds tensions with the dispersive representation, with the omega and phi widths, and with the resulting 3π contribution to the muon anomalous magnetic moment, reflected in an upward pull on C_A.","tokens_in":18307,"tokens_out":5496,"duration_ms":52510,"significance":"If correct, this is a valuable new test of the WZW anomaly: it uses only e+e−→3π data and exploits the dispersive sum rule to tie the anomaly to the omega and phi resonance region, making the extraction substantially more sensitive than low-energy data alone. The error breakdown into statistical, conformal-truncation, and rho-omega phase components is transparent, and the careful treatment of bin reweighting and iterative fitting is a strength. The paper also explicitly identifies tensions with Belle II data instead of forcing a combined result. Importantly, there is no circularity: the sum rule is a dispersion identity, and C_A is a fitted output compared against an external chiral benchmark. The main limitation is that the quoted precision depends on the convergence and completeness of the conformal-polynomial representation, which the fit-quality and truncation tests do not yet fully establish.","major_comments":[{"comment":"The quoted truncation uncertainty of 0.048 is essentially the full shift between p_conf=3 and p_conf=4 for delta_epsilon=0, and the sequence 1.038(11), 1.028(13), 0.980(16) is monotonic in p_conf. Figure 1 shows the result continues to move at p_conf=5, and the text states that for larger p_conf the fits become increasingly unstable. A shift equal to the quoted systematic is not by itself a demonstration of convergence. Please provide a sharper justification for choosing p_conf=3 as the central value and for treating the set {2,3,4} as an uncertainty envelope, or an independent stability check such as a prior on the higher coefficients, a data-driven order-selection criterion, or a physics-motivated asymptotic constraint on Im C_p.","section":"Sec. 3, Table 3, Eq. (3.3)"},{"comment":"Even after scale-factor inflation, the global-fit p-values are between 1e-5 and 4e-5 (chi2/dof about 1.31-1.37), so the fitted representation does not describe the data statistically. Because C_A is extracted as an integral over the fitted imaginary part, including the conformal polynomial up to arbitrarily large s', a model that fails at this level could bias the integral in a way not captured by the quoted uncertainties. Please quantify the impact of the model incompleteness, for example by showing results from single-dataset fits, from omitting the highest-energy BaBar points, or from adding an additional high-energy contribution or an alternative asymptotic form, and report the resulting shift in C_A/C_WZW_A.","section":"Tables 1 and 3; Eq. (2.11)"},{"comment":"The Belle II fits have p-values around 1e-8 and show significant tensions in the omega and phi widths and in a_mu. This is consistent with the model-incompleteness concern raised above and should be discussed not only as a property of the Belle II data set but also as a systematic check on the reliability of the global-fit extraction, especially since the upward pull in C_A from Belle II is of similar size to the global-fit total uncertainty.","section":"Sec. 4, Tables 4-6"}],"minor_comments":[{"comment":"The notation sthr = M^2_pi0 for the omega->pi0 gamma threshold would be clearer if written as M_{pi^0}^2 with an explicit statement that this is the square of the neutral-pion mass.","section":"Sec. 2, after Eq. (2.8)"},{"comment":"The error decomposition in the first column would be easier to use if the table footnote stated explicitly that the second error is the maximal variation over p_conf in {2,4} and the third error is the variation between delta_epsilon=0 and delta_epsilon=3.5 degrees.","section":"Table 2"},{"comment":"The phrase 'thereof 174 data points' should read 'of which 174 data points'.","section":"Sec. 4"},{"comment":"Reference [113] is a private communication; if a public Belle II auxiliary note or release description exists, citing it would make the singular-value cut reproducible.","section":"References"},{"comment":"Adding a horizontal line at C_A/C_WZW_A = 1 would make the comparison with the WZW prediction more direct.","section":"Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The paper builds on the authors' own previous formalism, and the novelty is the free-C_A fit, which is appropriate. The main risk is that the abstract's 'agreement at the level of 5%' could be over-read given the poor fit quality; the revision should make the conditional nature of the extraction more explicit and add the requested stability tests."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline: this is the first extraction of F3π from e+e−→3π data, and it is a sensible consistency check for the dispersive machinery used in muon g−2 work. The central result, F3π = 33.1(1.7) GeV^-3, agrees with the WZW prediction at about the 5% level, but that precision is softer than it appears. The quoted error is dominated by the truncation of the conformal polynomial, and the sequence CA/CWZW_A = 1.038(11), 1.028(13), 0.980(16) as pconf goes 2→3→4 shows the central value moving by the full quoted truncation error in the last step. The paper itself notes the fits become increasingly unstable for larger pconf, so the expansion has not demonstrably converged. Because the sum rule (2.11) makes the extracted anomaly an integral over the imaginary part, and the pconf contributions enter above 1 GeV^2 where the data only constrain up to 1.8 GeV, the 5% claim does depend on an unverified high-energy extrapolation. The stress-test note lands.\n\nThat said, credit where it is due. The authors are upfront about the poor global fit quality (p-value around 1e-5 even after scale-factor inflation) and they include the truncation spread as a systematic error rather than hiding it. The Belle II comparison is genuinely useful: those data pull Γω and Γϕ down by 3.2σ and 2.6σ and push the anomaly upward, which is either a data-set issue or a sign the representation is missing something. The authors frame it as a tension to monitor, which is honest.\n\nThe circularity concern is mostly a non-issue. F3π is a fitted parameter, the sum rule is a dispersion identity, and the agreement with WZW is an external benchmark. The same group's earlier papers supply the formalism, but that is not circular.\n\nThe main weakness is that the central value's stability is not established. The 0.048 shift from pconf 3 to 4 equals the quoted truncation error, so the 5% total uncertainty is an envelope, not a convergence statement. A serious referee should ask for a more principled truncation criterion, such as a Bayesian model average or a check against an alternative parameterization, rather than just the spread across pconf = 2, 3, 4.\n\nBottom line: this is a useful, honest paper and it deserves proper peer review, not a desk rejection. Send it out with a request to confront the convergence issue head-on and to soften the language about \"5% precision\" if the systematic envelope is dominated by an unstable expansion. Conditional acceptance is the right call.\n\nBest,\n[You]","headline":"First extraction of the chiral anomaly from e+e−→3π data is a useful consistency check, but the 5% precision claim is softer than it looks because the conformal-polynomial expansion has not demonstrably converged.","tokens_in":18755,"tokens_out":2481,"would_cite":true,"duration_ms":23193,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that a dispersive fit to $e^+e^-\\to 3\\pi$ data extracts the chiral anomaly strength $F_{3\\pi}$ at about 5% precision, in agreement with the Wess–Zumino–Witten prediction.","keywords":["chiral anomaly","Wess–Zumino–Witten","e+e−→3π","dispersive representation","Khuri–Treiman formalism","pion decay constant","hadronic vacuum polarization","muon g−2"],"falsifier":"Evaluate the sum rule $\\alpha_A=\\frac{1}{\\pi}\\int ds'\\,\\mathrm{Im}\\,a(s')/s'$ directly from the measured cross section over the full energy range, without assuming the fitted parameterization, and compare the resulting $\\alpha_A$ with $C_A/3$ from the fit; a disagreement larger than the quoted $5\\%$ uncertainty would falsify the extraction, as would a future high-statistics Belle II measurement whose fitted $C_A/C_A^{\\mathrm{WZW}}$ excludes $1.028$ by more than the combined errors.","tokens_in":17719,"feed_emoji":"⚛️","tokens_out":8931,"duration_ms":72477,"temperature":0.7,"pith_summary":"The paper aims to convert measurements of $e^+e^-\\to 3\\pi$ into a precision test of the chiral (Wess–Zumino–Witten) anomaly, which predicts the strength $F_{3\\pi}$ of the $\\gamma^*\\to 3\\pi$ coupling from the pion decay constant. Using a dispersive representation in which the anomaly enters as a subtraction constant tied through a sum rule to the integral of the amplitude's imaginary part, the authors fit cross-section data from SND, CMD-2, and BaBar and obtain $F_{3\\pi}=33.1(1.7)\\,\\mathrm{GeV}^{-3}$, agreeing with the chiral prediction at the $5\\%$ level. If the extraction is sound, it validates the anomaly prediction at a precision comparable to existing low-energy measurements, and it supports the common practice of imposing the anomaly as a constraint in dispersive calculations of hadronic processes.","feed_headline":"e+e−→3π data measure the chiral anomaly to 5 percent precision","feed_subtitle":"A dispersive fit gives F3π = 33.1(1.7) GeV−3, matching the Wess–Zumino–Witten prediction within errors.","key_machinery":"The machinery is a dispersive (Khuri–Treiman) representation of the $\\gamma^*\\to 3\\pi$ amplitude, writing the scalar function $F(s,t,u;q^2)$ in terms of partial-wave amplitudes and a subtraction function $a(q^2)$. The anomaly enters through the subtraction constant $\\alpha_A=C_A/3$ in $a(q^2)$, and the load-bearing identity is the sum rule (2.11), which is imposed at every fit step and makes the $\\omega$, $\\phi$, and $\\omega'$ resonances plus a truncated conformal polynomial carry the anomaly information. The formalism includes $\\rho$–$\\omega$ mixing through a correction factor and dispersively improved Breit–Wigner parameterizations for the resonances.","core_discovery":"The central claim is that the chiral anomaly can be extracted from $e^+e^-\\to 3\\pi$ data as a global fit parameter rather than imposed as an input. In the dispersive framework the subtraction constant $\\alpha_A=C_A/3$ is fixed by the sum rule $\\alpha_A=\\frac{1}{\\pi}\\int_{s_{\\mathrm{thr}}}^{\\infty} ds'\\,\\mathrm{Im}\\,a(s')/s'$, so the anomaly becomes a property of the whole measured cross section rather than only its low-energy threshold. Fitting $C_A$ to SND, CMD-2, and BaBar data yields $C_A/C_A^{\\mathrm{WZW}}=1.028(15)(48)(17)[53]$, corresponding to $F_{3\\pi}=33.1(1.7)\\,\\mathrm{GeV}^{-3}$, consistent with the WZW prediction within $5\\%$. Fits to recent Belle II data instead show tensions with the dispersive constraints, lower $\\omega$ and $\\phi$ widths, and an upward pull in both $a_\\mu^{3\\pi}$ and the extracted anomaly, which the paper interprets as a problem in the interpretation of that data set.","pith_inferences":["A testable extension would be a data-driven evaluation of the sum rule directly from binned cross sections above 1 GeV, bypassing the conformal polynomial; the paper imposes the sum rule but does not compare it with an unparameterized integral.","If the Belle II tensions are resolved by future data with proper bin weighting, a combined fit could push the anomaly extraction below 5%, making $e^+e^-\\to 3\\pi$ the sharpest single observable for $F_{3\\pi}$.","The same formalism applies to $\\eta\\to 3\\pi$ and $\\eta'\\to 3\\pi$, where the anomaly normalization enters analogously, offering a cross-channel check of the WZW prediction not performed in this paper."],"forward_implications":["The WZW prediction for $\\gamma\\to 3\\pi$ is supported at $5\\%$ precision, making $F_{3\\pi}$ a reliable input for dispersive hadronic light-by-light and $e^+e^-\\to 3\\pi$ cross-section calculations entering the muon $g-2$.","The anomaly can be left as a free parameter in future $e^+e^-\\to 3\\pi$ fits, turning every new data set into an independent consistency check on the chiral prediction.","The global fit determines $\\omega$ and $\\phi$ masses and widths consistent with literature values, while Belle II data pull the widths down by $3.2\\sigma$ and $2.6\\sigma$, showing that the dispersive fit can expose tensions inside a data set.","The upward pull in $C_A$ from Belle II data tracks the upward pull in $a_\\mu^{3\\pi}$, so resolving the Belle II tensions will sharpen both the anomaly test and the hadronic-vacuum-polarization estimate for the muon's anomalous magnetic moment."],"supporting_citations":[{"why":"Supplies the dispersive Khuri–Treiman representation of $\\gamma^*\\to 3\\pi$, including the sum rule that ties the anomaly constant to the imaginary part of the amplitude.","marker":"[47]"},{"why":"Provides the extended formalism with $\\rho$–$\\omega$ mixing and the global fit whose conventions and data selection this paper adopts.","marker":"[75]"},{"why":"BaBar ISR cross-section data, a key input of the global fit, for which bin reweighting is essential.","marker":"[96]"},{"why":"Belle II data whose fit produces the tensions in $\\omega$/ $\\phi$ widths, $a_\\mu^{3\\pi}$, and the upward anomaly pull.","marker":"[97]"},{"why":"SND data (series [88–91]) used in the global fit.","marker":"[88]"},{"why":"CMD-2 data (series [92–95]) used in the global fit.","marker":"[92]"},{"why":"Derives the quark-mass correction $C_A=1.066(10)F_{3\\pi}$ used to convert the fitted $C_A$ to $F_{3\\pi}$.","marker":"[19]"},{"why":"Suggested the dispersive strategy of leveraging the $\\rho(770)$ peak for anomaly extraction, extended here to $e^+e^-\\to 3\\pi$.","marker":"[27]"}],"fun_headline_variants":["Chiral anomaly pinned to 5% from e+e−→3π data","Dispersive fit gives chiral anomaly within 5% of prediction","e+e−→3π global fit sets chiral anomaly at 5% level","Chiral anomaly from e+e−→3π matches theory to 5%","Belle II data show tension in chiral anomaly extraction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The extraction depends on the assumption that the fitted $\\omega$, $\\phi$, and $\\omega'$ resonances plus a truncated conformal polynomial capture all of the dispersive strength in the measured region, so any missing high-energy contribution or wrong energy dependence of the polynomial would shift the extracted anomaly value.","fun_headline_variants_meta":{"raw":{"variants":["Chiral anomaly pinned to 5% from e+e−→3π data","Dispersive fit gives chiral anomaly within 5% of prediction","e+e−→3π global fit sets chiral anomaly at 5% level","Chiral anomaly from e+e−→3π matches theory to 5%","Belle II data show tension in chiral anomaly extraction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001432,"raw_usage":{"total_tokens":5796,"prompt_tokens":989,"completion_tokens":4807,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":605,"completion_tokens_details":{"reasoning_tokens":4710}},"tokens_in":605,"tokens_out":4807,"duration_ms":30399,"temperature":1.0,"reasoning_tokens":4710,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:59:20.281622+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the sum rule $\\alpha_A=\\frac{1}{\\pi}\\int ds'\\,\\mathrm{Im}\\,a(s')/s'$ directly from the measured cross section over the full energy range, without assuming the fitted parameterization, and compare the resulting $\\alpha_A$ with $C_A/3$ from the fit; a disagreement larger than the quoted $5\\%$ uncertainty would falsify the extraction, as would a future high-statistics Belle II measurement whose fitted $C_A/C_A^{\\mathrm{WZW}}$ excludes $1.028$ by more than the combined errors.","supporting_citations":[],"review_version":1}