{"id":"f79c103c-e86e-4226-96aa-4f064ff4960a","arxiv_id":"2504.13789","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A data-only extraction of the charged versus neutral rho mass and width differences gives a total isospin-breaking correction of -(12.2 +/- 3.4) x 10^-10 to the tau-based hadronic vacuum polarization contribution to the muon g-2.","lead":"This paper measures the mass and width of the neutral and charged rho mesons directly from the shapes of e+e- and tau decay data, instead of using theoretical estimates. The resulting isospin-breaking correction shifts the tau-based prediction of the muon magnetic anomaly by about 2.7 x 10^-10 and makes it less certain.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Tail isospin breaking is only partially tested: freeing ρ′/ρ′′ amplitudes in §5.3 leaves their masses and widths fixed to BABAR e+e− values, so a physical charged/neutral splitting of the excited states would bias Δmρ and ΔΓρ.","rationale":"The reader's weakest assumption points to the high-mass tail being taken from BABAR e+e− data and applied to τ data; my concern is the same assumption, sharpened to a specific untested direction. Section 5.3 is a genuine and useful test, but it varies only the relative amplitudes of the excited states, not their masses or widths. Since the ρ′ amplitude is large enough that ρ−ρ′ interference dominates the tail under the ρ peak (Figure 6), a small isospin splitting in the excited-state masses would be absorbed by the fitted ρ parameters and would bias the very differences the paper aims to measure. The 5% oscillations seen in the Belle and CMD-3 comparisons (Figure 1) show that the fixed-tail assumption is not exact at the precision needed. The proposed synthetic shift test would directly quantify this bias and decide whether the quoted 0.38 MeV/0.25 MeV systematics are sufficient. The paper otherwise contains strong independent support: the normalization-decoupling check in Figure 4 is clever, the GS-vs-KS comparison is honestly quantified, and the GEM(s) sensitivity is explicitly addressed. These do not resolve the tail-IB question, but they do show that the authors are attentive to model dependence. For that reason I do not recommend changing the verdict from CONDITIONAL; the paper remains acceptable in principle, with the tail-IB direction identified as the point that a revision should close.","tokens_in":16608,"tokens_out":7287,"duration_ms":71589,"concrete_test":"Refit the τ spectra of §3 with the ρ′ mass shifted by +2 MeV and then by −2 MeV, and likewise for Γρ′; repeat the exercise with ρ′′, keeping all other BABAR tail parameters fixed and refitting mρ and Γρ. Record the induced shifts in mρ± and Γρ±. Separately, in the Belle fit above 0.9 GeV, free the ρ′ mass, width, and amplitude simultaneously and compare the fitted values with the BABAR e+e− values. If the induced shifts exceed roughly 0.3 MeV for the mass or 0.5 MeV for the width, the §5.3 tail systematic is underestimated and the extracted Δmρ/ΔΓρ, and hence the IB correction, should carry an additional tail-IB uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that a data-based extraction of Δmρ and ΔΓρ can replace the theoretical IB correction. That extraction is trustworthy only if the difference between the e+e− and τ fit shapes is actually generated by the ρ ground state, not by the higher ρ states whose tails are imposed on the fit. Section 3 fixes the higher-mass contributions to the BABAR e+e− fit parameters. Section 5.3 checks the tail by freeing only the relative complex amplitudes of ρ′ and ρ′′ in τ fits and then quotes 0.38 MeV (mass) and 0.25 MeV (width) as systematic uncertainties. This test does not vary the masses or widths of the excited states. Isospin breaking in the tail could equally appear as δm_{ρ′}, δΓ_{ρ′}, δm_{ρ′′}, δΓ_{ρ′′}; with |c_{ρ′}|≈0.15 and pronounced ρ−ρ′ interference under the ρ peak (Figure 6), even a 1–2 MeV shift of m_{ρ′} generates a slope in the fit region that the fitted ρ mass and width would absorb. The empirical support for the fixed-tail assumption is also limited: Figure 1 shows Belle and CMD-3 data oscillating at the 5% level against the BABAR-based tail. The quoted §5.3 systematics therefore cover only one slice of the possible tail-IB space. The final ΔΓρ has a 1.04 MeV uncertainty and its central value differs from the theoretical estimate by 1.3σ; a tail-induced bias of order 0.5 MeV in Γρ or 0.3 MeV in mρ would be material, so this is the load-bearing point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes replacing the theoretical estimate of the isospin-breaking (IB) correction between the e+e- and tau two-pion spectral functions with a data-based determination. The authors fit the rho resonance parameters in each e+e- and tau dataset with a Gounaris-Sakurai form factor, leaving the normalization free to decouple the fitted mass and width from the absolute spectral-function scale. The high-mass tail is fixed from the wide BABAR e+e- fit. From the per-experiment fits they derive Delta m_rho = (-0.30 +/- 0.53) MeV and Delta Gamma_rho = (-0.58 +/- 1.04) MeV, combine them with the other IB contributions, and obtain a total IB correction of -(12.2 +/- 3.4) x 10^-10 for the tau-based HVP prediction of a_mu. The paper includes consistency checks for normalization decoupling, line-shape parametrization, high-mass tail sensitivity, and the GEM correction, and compares the resulting tau-based a_mu with e+e- results and the Fermilab measurement.","tokens_in":16984,"tokens_out":5072,"duration_ms":49398,"significance":"If the central claim holds, this is a valuable methodological step: it offers a data-driven alternative to the model-based pion form-factor IB correction that has long been one of the weaker points in tau-based HVP estimates. The paper is careful in several respects: the normalization decoupling is explicitly tested via the correlation between fitted normalizations and a_mu integrals (Figure 4), the quoted uncertainties are honest and conservative, and the consistency of the extracted rho parameters across experiments is examined in detail. The final result is consistent with the previous theory-driven correction, although with larger uncertainty, and it provides an independent cross-check in a field where independent cross-checks are scarce. The main weakness is that the high-mass tail is imposed on the tau fits from BABAR e+e- data and the sensitivity test only partially covers the associated isospin-breaking uncertainty; this point is load-bearing for the central claim and needs additional quantitative support.","major_comments":[{"comment":"The tail-isospin-breaking test is incomplete: freeing only the relative complex amplitudes of rho' and rho'' while keeping their masses and widths fixed to the BABAR e+e- values does not explore isospin breaking in the excited-state masses and widths. A physical charged/neutral splitting of rho' or rho'' would be absorbed into the fitted rho mass and width through the pronounced rho-rho' interference shown in Figure 6, and would bias the quoted Delta m_rho and Delta Gamma_rho. The shifts quoted in Section 5.3 (0.38 MeV in m_rho and -0.25 MeV in Gamma_rho) therefore cover only one slice of the allowed parameter space. Since the final uncertainties are 0.53 MeV (mass) and 1.04 MeV (width), an unquantified tail-IB bias of order 0.3-0.5 MeV is material. The authors should extend the study by varying delta m_rho', delta Gamma_rho', delta m_rho'', and delta Gamma_rho'' (or otherwise bounding them using the tau data), and should propagate the resulting shifts as an additional systematic uncertainty.","section":"Section 5.3, Eqs. (19) and (20)"},{"comment":"The empirical support for the fixed-tail assumption is weaker than the text suggests. In Figure 1, both CMD-3 and Belle data show deviations at the 5% level relative to the BABAR-tail extrapolation in the fitted region, and these residual oscillations are not converted into a quantitative uncertainty on the extracted rho parameters. The paper should translate these observed deviations into a systematic uncertainty, for example by refitting with the tail parameters or amplitudes varied within the envelope suggested by the data/fit ratios, rather than relying only on the free-amplitude check in Section 5.3.","section":"Section 3 and Figure 1"},{"comment":"The conclusion that the data-based method 'removes the issue related to the reliability' of the theoretical radiative-width calculation is stronger than the results support: the data-based Delta Gamma_rho has a 1.04 MeV uncertainty and is consistent with the theoretical prediction of 0.76 +/- 0.18 MeV at only 1.3 sigma, so it cannot validate or invalidate that calculation at the claimed accuracy. The text should state this limitation explicitly and frame the data-based value as a conservative replacement rather than as a resolution of the theoretical uncertainty.","section":"Section 9 and Table 3"}],"minor_comments":[{"comment":"The form factors F_0(s) and F_-(s) are introduced only verbally; they should be defined explicitly in the text or in the equation, since the sign convention in Eq. (2) is central to the IB correction.","section":"Section 2, Eq. (2)"},{"comment":"The sentence 'The small GEM shape correction is the only explicit model-dependence introduced in this procedure' appears to contradict the earlier statement in the same section that FSR is also model-dependent and evaluated with scalar QED; please clarify which model dependences are considered 'explicit' here.","section":"Section 2"},{"comment":"There are typographical and spacing issues, for example 'tau ande+e-' in the abstract and the italic/spacing treatment of quantities such as a_tau_mu and a_ee_mu; a careful proofread is recommended.","section":"Abstract and Introduction"},{"comment":"Figure 6 would be easier to interpret if the interference contributions were shown separately or with a clear normalization, since the statement that rho-rho' interference dominates the mass shift is important for the tail-IB discussion.","section":"Section 5.3, Figure 6"},{"comment":"The quoted uncertainties would be more transparent if the systematic contributions (normalization decoupling, fit-range, binning, line-shape, tail, GEM) were itemized in a table or text list, rather than appearing only as combined values in Table 3.","section":"Section 6 and Table 3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a natural extension of the authors' earlier work [8] and is well within the scope of a high-energy physics journal. The central methodological idea is sound and the consistency checks are generally convincing. The main issue is that the high-mass-tail isospin-breaking uncertainty is not fully explored; this is fixable with additional fits and should be addressed before acceptance. I would not want to see the paper rejected on this point, since the authors are in the best position to run the extended sensitivity study."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a real look. The genuinely new piece is the extraction of Delta m_rho and Delta Gamma_rho from fits with free normalization, so the shape parameters are decoupled from the absolute scale. Figure 4 is the key evidence that the decoupling works: KLOE and CMD-3 differ by 5 sigma in a_mu but their fitted normalizations track that difference, leaving the rho parameters consistent. That is a solid idea, carefully executed. The KS/GS comparison and the GEM check are also honest, and the final uncertainty reflects real limitations.\n\nThe soft spots are real but not fatal. The KLOE mass is excluded because of a slope that the fit cannot absorb; the paper explains why, but it is a post-hoc choice and the averaging should probably show the result with KLOE included as a robustness test. More important is the tail assumption. Section 5.3 frees the rho' and rho'' amplitudes in tau fits, but keeps their masses and widths fixed to the BABAR e+e- values. That test covers one slice of the possible isospin-breaking space. If the excited states themselves have charged/neutral mass or width splittings, the fitted ground-state rho parameters will absorb some of that, and the quoted 0.38 MeV and 0.25 MeV systematics do not include it. The 5% oscillations between the BABAR tail and CMD-3/Belle data in Figure 1 suggest the tail is not perfectly universal. So the central result is conditional: it is a genuine data-based estimate under an assumption that is plausible but not fully tested.\n\nThe paper is also slightly oversold as 'based only on data': the GEM shape correction and the GS parametrization are model-dependent, and the high-mass tail is entirely fixed from one e+e- experiment. That does not invalidate the method, but the 'data-based' claim should be tempered in the abstract.\n\nCitation pattern is fine; the authors cite the relevant theory estimates and their own previous work appropriately. One input comes from a private communication (Belle systematics), which is worth flagging but not disqualifying.\n\nMy take: this deserves a serious referee. The method is useful, the numerical result is honest about its precision, and the soft spots are addressable in revision. I would bring it to a reading group and would probably cite it for the free-normalization technique, even if I do not adopt the central numbers as-is.","headline":"First data-based rho mass/width differences for tau vs e+e-; the method is sound but the tail-isospin assumption is only partially stress-tested.","tokens_in":17544,"tokens_out":949,"would_cite":true,"duration_ms":11487,"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":"Data-based fits of the rho resonance shape determine the isospin-breaking correction between tau and e+e- two-pion spectral functions, replacing the previous theory-driven estimate.","keywords":["isospin breaking","pion form factor","tau spectral function","e+e- annihilation","muon g-2","hadronic vacuum polarization","Gounaris-Sakurai","rho meson"],"falsifier":"Fit the $\\tau$ two-pion spectral function over its full kinematic range with the $\\rho'$ and $\\rho''$ complex amplitudes left free; if the fitted amplitudes differ from the fixed $e^+e^-$ values by more than the amounts that shift $m_\\rho$ by 0.38 MeV or $\\Gamma_\\rho$ by 0.25 MeV, the fixed-tail assumption fails and the central isospin-breaking correction is biased.","tokens_in":16401,"feed_emoji":"⚛️","tokens_out":13587,"duration_ms":103339,"temperature":0.7,"pith_summary":"The paper replaces a theory-based estimate with a data-based determination of the isospin-breaking (IB) correction that connects the two-pion spectral functions measured in tau decays and in e+e- annihilation. The authors fit the rho resonance line shape to all precise e+e- and tau datasets with the overall normalization left free, so the fitted rho mass and width are decoupled from the absolute scale of each measurement. The resulting differences, $\\Delta m_\\rho = (-0.30\\pm 0.53)$ MeV and $\\Delta \\Gamma_\\rho = (-0.58\\pm 1.04)$ MeV, feed into an updated total IB correction of $-(12.2\\pm 3.4)\\times 10^{-10}$ for the tau-based hadronic vacuum polarization contribution to the muon g-2. This value agrees with earlier theory-based corrections but carries a larger uncertainty, and it removes the reliance on a model-dependent radiative-width calculation. If correct, the method gives an independent, fully data-driven route to correct tau spectral functions before they are used in dispersion relations.","feed_headline":"Rho mass split measured from data, replacing a model estimate","feed_subtitle":"Shape fits yield the rho mass split and update the tau-based muon g-2 correction to -12.2 ± 3.4 ×10^-10.","key_machinery":"The engine of the method is the Gounaris-Sakurai (GS) parametrization of the $\\rho$ resonance, a Breit-Wigner line shape with an energy-dependent width and analyticity-preserving auxiliary functions, fitted to each dataset with a free overall normalization. Freeing the normalization decouples the fitted mass and width from the absolute scale of the measurement, so the parameters carry shape information only and are not circularly tied to the $a_\\mu$ integral. The higher-mass contributions $\\rho'$, $\\rho''$, and $\\rho'''$ are fixed from a wide $e^+e^-$ fit rather than fitted per experiment, and $e^+e^-$ fits add four parameters for $\\rho$–$\\omega$ interference; the fit range is restricted below 0.9 GeV for all datasets.","core_discovery":"The paper's central claim is that the isospin-breaking difference between the charged and neutral $\\rho$ resonance parameters can be extracted directly from the shapes of the $e^+e^-$ and $\\tau$ two-pion spectral functions. Using the Gounaris-Sakurai parametrization with a free normalization for every dataset, the authors obtain averaged values $m_{\\rho^0}=(774.93\\pm 0.12)$ MeV and $\\Gamma_{\\rho^0}=(148.74\\pm 0.19)$ MeV from $e^+e^-$ data and $m_{\\rho^\\pm}=(775.23\\pm 0.52)$ MeV and $\\Gamma_{\\rho^\\pm}=(149.32\\pm 1.02)$ MeV from $\\tau$ data, giving $\\Delta m_\\rho = (-0.30\\pm 0.53)$ MeV and $\\Delta\\Gamma_\\rho = (-0.58\\pm 1.04)$ MeV. These are then used to compute the pion form factor part of the total isospin-breaking correction, which, together with the separately fitted $\\rho$–$\\omega$ interference, yields a total correction $-(12.2\\pm 3.4)\\times 10^{-10}$. The authors argue that this data-based determination is consistent with earlier theoretical estimates but replaces the model-dependent radiative width calculation, and they validate the shape-only extraction by showing that each experiment's fitted normalization tracks its dispersion-integral value.","pith_inferences":["Editorial inference: the same shape-versus-normalization decoupling could be extended to other hadronic channels (for instance four pions) to build fully data-driven isospin-breaking corrections for the rest of the hadronic vacuum polarization integral.","Editorial inference: if isospin breaking in the high-mass tail is larger than assumed, the quoted 0.38 MeV mass systematic would need to scale up; a dedicated $\\tau$ high-mass measurement with free $\\rho'$ amplitudes would settle this.","Editorial inference: the slope observed in one $e^+e^-$ dataset relative to the others points to a second-order shape difference rather than pure normalization; understanding its origin could resolve part of the $e^+e^-$ tension.","Editorial inference: the normalization-to-integral ratio test could be reused as a general cross-check of $e^+e^-$ averaging procedures, since it cleanly separates scale effects from shape effects."],"forward_implications":["The total isospin-breaking correction becomes $-(12.2\\pm 3.4)\\times 10^{-10}$, consistent with the earlier theory-based values but with roughly double the uncertainty.","With this correction, the $\\tau$-based $a_\\mu$ prediction moves closer to the most precise $e^+e^-$ results and sits $1.8\\sigma$ from the direct muon g-2 measurement.","The limiting factor is the precision of existing $\\tau$ data; future higher-statistics $\\tau$ datasets and the large $J/\\psi\\to\\rho\\pi$ sample are expected to reduce the uncertainties.","Once normalization differences are removed, the $\\rho$ parameters from all $e^+e^-$ experiments are consistent except for a mass slope in one dataset, indicating that part of the known $e^+e^-$ tension is a pure scale effect.","The data-based width difference carries an uncertainty about six times larger than the theoretical prediction, so it cannot yet check the radiative-width calculation at its claimed accuracy."],"supporting_citations":[{"why":"Supplies the Gounaris-Sakurai parametrization used for every rho shape fit.","marker":"[27]"},{"why":"Provides the wide e+e- pion form factor fit whose higher-mass resonance parameters are fixed in all other fits.","marker":"[28]"},{"why":"Previous analysis that supplies the comparison framework and non-two-pion inputs for the final muon g-2 predictions.","marker":"[8]"},{"why":"Tau spectral function dataset and combined analysis used in the tau fits and uncertainty treatment.","marker":"[10]"},{"why":"The model-dependent radiative correction whose energy dependence is applied before the shape fits.","marker":"[24]"},{"why":"The radiative rho width calculation that the data-based form factor correction replaces.","marker":"[26]"},{"why":"The most precise e+e- scan dataset in the rho region, dominating the e+e- average.","marker":"[36]"},{"why":"Direct muon g-2 measurement used as the reference in the final comparison.","marker":"[47]"}],"fun_headline_variants":["Data-based rho mass split replaces model estimate","Rho mass split measured from spectral shapes","Tau and e+e- data pin rho mass split","Data-driven rho split updates muon g-2","Rho mass split from data, not models"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the high-mass tail of the pion form factor, taken from a wide $e^+e^-$ fit, is identical for $\\tau$ and $e^+e^-$ data, so isospin breaking in the $\\rho'$ and higher resonances is neglected and would bias the fitted mass and width differences if it were significant.","fun_headline_variants_meta":{"raw":{"variants":["Data-based rho mass split replaces model estimate","Rho mass split measured from spectral shapes","Tau and e+e- data pin rho mass split","Data-driven rho split updates muon g-2","Rho mass split from data, not models"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000336,"raw_usage":{"total_tokens":1881,"prompt_tokens":986,"completion_tokens":895,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":602,"completion_tokens_details":{"reasoning_tokens":834}},"tokens_in":602,"tokens_out":895,"duration_ms":7800,"temperature":1.0,"reasoning_tokens":834,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:59:36.396401+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fit the $\\tau$ two-pion spectral function over its full kinematic range with the $\\rho'$ and $\\rho''$ complex amplitudes left free; if the fitted amplitudes differ from the fixed $e^+e^-$ values by more than the amounts that shift $m_\\rho$ by 0.38 MeV or $\\Gamma_\\rho$ by 0.25 MeV, the fixed-tail assumption fails and the central isospin-breaking correction is biased.","supporting_citations":[{"cited_title":"Update of the ALEPH non-strange spectral functions from hadronic τ decays","cited_arxiv_id":null,"evidence_quote":"Tau spectral function dataset and combined analysis used in the tau fits and uncertainty treatment."}],"review_version":1}