{"id":"44f64d5d-509f-433b-a318-f9fa0d52b95f","arxiv_id":"2412.11989","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":10,"one_line_summary":"ATLAS measures R_tau/e = 0.975 +/- 0.012 (stat) +/- 0.020 (syst) for W boson decays to tau versus electron, consistent with lepton flavour universality.","lead":"ATLAS used 140 inverse femtobarns of LHC proton-proton collisions to measure how often a W boson decays to a tau lepton versus an electron, using top-quark decays as the W source. The ratio is 0.975 with a total uncertainty of 0.024, consistent with the Standard Model, and helps settle a two-standard-deviation tension in older LEP results.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fake-electron corrections from the same-sign control sample are transported to the opposite-sign signal region using simulation only; a charge-sign-dependent bias in this transfer would shift Rτ/e by more than the assigned background systematic.","rationale":"The reader correctly identified the same-sign to opposite-sign fake-electron extrapolation as the weakest assumption. I agree that this is the most load-bearing point because the correction factors are fitted in a sample with 97% fake purity at low pT, while the signal region has only about 10% fakes; the transfer between these very different compositions is performed purely with simulation. The paper does assign a background systematic (0.005), but the systematic is derived from generator comparisons and photon-conversion fraction variations, not from an independent OS data control region. Given that a modest bias in the OS fake rate could shift R_tau/e by an amount comparable to the quoted total uncertainty, the precision claim depends on an untested assumption. This concern does not invalidate the consistency with the Standard Model, but it affects the reliability of the quoted systematic uncertainty and hence the claim of improved world-average precision. The proposed m_ll < 15 GeV closure test is a concrete, low-cost check that would settle whether the extrapolation is biased. I recommend CONDITIONAL acceptance: the measurement is credible, but it should be accepted only after the authors provide this (or an equivalent) data-driven validation of the SS-to-OS transfer.","tokens_in":58753,"tokens_out":8618,"duration_ms":82689,"concrete_test":"Use the currently excluded m_ll < 15 GeV opposite-sign region (removed in Section 4 to suppress fakes) as a fake-enriched control sample. After applying the same-sign-derived correction factors and the MC-based SS-to-OS transfer, compare data and prediction in bins of probe pT and |d0|. If the closure is not within the assigned background systematic, the SS-to-OS extrapolation is biased and R_tau/e should be re-derived with an enlarged uncertainty or a direct OS fake estimate.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The most load-bearing assumption is in Section 6: fake-electron correction factors (C_HAD, C_PH, C_PR) are fitted in a same-sign (SS) control sample and then applied to the opposite-sign (OS) signal region, with the SS-to-OS transfer factor taken entirely from simulation. The paper states: 'The extrapolation of the correction factor from the same-sign to opposite-sign sample is evaluated from simulated events.' The SS sample is fake-dominated (97% at pT < 10 GeV, versus roughly 10% in the OS signal region), so the fitted corrections are anchored to a sample with very different fake composition. A bias in the MC transfer—for example, in the charge-asymmetric fraction of hadron-decay fakes or photon conversions—would shift the OS fake yield and, because fakes and tau-to-e electrons both populate large |d0| and low pT, would bias R_tau/e. The assigned 'Background estimation' systematic (0.005 in Table 4) is derived from generator comparisons (Pythia vs Herwig) and the photon-conversion fraction, but it does not include a data-driven test of the SS-to-OS extrapolation itself. Quantitatively, in the 7-10 GeV bin the fake yield is about 10% of the total; a 20% error in the OS fake rate corresponds to roughly 2% of the tau signal in that bin, comparable to the quoted total uncertainty. No opposite-sign control region enriched in fakes is presented to close this loop.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The ATLAS Collaboration presents the first measurement of R_tau/e = B(W -> tau nu)/B(W -> e nu) using top-quark decays, with 140 fb^-1 of pp collision data at sqrt(s)=13 TeV. Events are selected with one tag lepton and one probe electron, and the two W-decay categories are separated using the electron transverse impact parameter and pT distributions in a binned profile likelihood fit. The measured value is R_tau/e = 0.975 +/- 0.012 (stat) +/- 0.020 (syst), consistent with the Standard Model expectation of lepton flavour universality. The analysis relies on data-driven d0 templates from Z->ee events, same-sign control samples to derive fake-electron correction factors, and a detailed systematic evaluation dominated by ttbar/Wt modelling, d0 calibration, and background estimation.","tokens_in":59007,"tokens_out":9674,"duration_ms":91094,"significance":"If correct, this measurement provides a new precision test of lepton flavour universality in on-shell W-boson decays, with an uncertainty comparable to the CMS result and the LEP combination, and it reduces the world-average uncertainty on R_tau/e. The statistical framework is clearly specified, with a binned likelihood and nuisance parameters, and the d0 calibration using a large Z->ee sample is a notable strength. The paper also ships a transparent systematic breakdown, including generator comparisons for the dominant modelling uncertainties. The principal weakness is the reliance on simulation for the same-sign to opposite-sign fake-electron extrapolation, which is not closed with a dedicated data control region.","major_comments":[{"comment":"The fake-electron correction factors (C_HAD, C_PH, C_PR) are fitted in a same-sign control sample and then transported to the opposite-sign signal region using a transfer factor evaluated entirely from simulation, as stated in Section 6: 'The extrapolation of the correction factor from the same-sign to opposite-sign sample is evaluated from simulated events.' The assigned 'Background estimation' systematic of 0.005 in Table 4 is derived from the same-sign fit, a Pythia/Herwig comparison, and a photon-conversion fraction variation, but it does not directly test the charge-sign-dependent transfer. In the 7-10 GeV pT bin, fakes are ~11% of the selected events (Table 1: 317 of 2770, compared to 1092 tau-signal events); a 20% bias in the OS fake rate would shift R_tau/e by roughly 0.03-0.05, which is comparable to or larger than the quoted total uncertainty of 0.024. I request either a data-driven closure test in an opposite-sign fake-enriched control region or an enlarged systematic that explicitly covers the SS-to-OS extrapolation, along with a discussion of how the current generator comparison constrains this source.","section":"Section 6 and Table 4"}],"minor_comments":[{"comment":"The pT-binned results show R_tau/e = 1.13 +/- 0.13 in the 7-10 GeV bin, which is the bin with the largest fake fraction and where the SS-to-OS extrapolation is most uncertain; a sentence interpreting this upward trend and its consistency with the combined result would be helpful.","section":"Section 8, Table 3"},{"comment":"The label 'Total systematical uncertainty' should read 'Total systematic uncertainty'.","section":"Table 4"},{"comment":"The likelihood uses Gaussian constraints for all nuisance parameters, including normalization-type uncertainties; for large pulls, log-normal constraints would be more appropriate, and a brief justification of the Gaussian choice would improve the statistical presentation.","section":"Eq. (3)"},{"comment":"The caption should state explicitly that the PDG average value shown does not include this new ATLAS measurement, otherwise a reader may assume the average already contains the result.","section":"Figure 8"},{"comment":"The statement 'It was verified that the separate treatment of electrons and muons from these sources had no statistically significant impact on the result' would benefit from a quantitative value (e.g., the shift in R_tau/e) so that the reader can assess the sensitivity.","section":"Section 6"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a standard high-quality ATLAS measurement, and the central result appears robust. The main concern is the same-sign to opposite-sign fake-electron extrapolation, which is a genuine limitation but not, in my view, a fatal flaw; the assigned background systematic may be optimistic. I recommend minor revision to address this with a closure test or a more detailed systematic justification, rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First ATLAS measurement of R_tau/e from top-quark decays, with precision comparable to the LEP combination and CMS. The result, 0.975 ± 0.024, is consistent with SM universality and helps resolve the old LEP tension. The analysis is careful: d0 templates are calibrated from Z->ee data, fake rates are fitted in a same-sign control sample, and the profile likelihood includes about 300 nuisance parameters. Post-fit distributions look good, with a global goodness-of-fit p-value of 87%. I find no circularity: R_tau/e is floated in the fit, not assumed.\n\nThe soft spot is the same-sign to opposite-sign extrapolation of fake-electron corrections. The paper states explicitly that this transfer is evaluated only from simulated events, with no opposite-sign control region enriched in fakes to test it. Given fakes are about 10% of the lowest-pT bin, a modest MC error in the charge-asymmetric fake fraction could shift R_tau/e by more than the assigned 0.005 background systematic. The collaboration does vary the photon-conversion fraction and the parton-shower model, which is a reasonable envelope, but it is not a full data-driven closure test. This is worth a referee question, not a rejection. The paper is transparent about the limitation, and the central value sits close to unity, so the conclusion is robust to plausible shifts.\n\nThe other limitations are standard for a large collider experiment: no public code or analysis software, and the internal detector-calibration chain cannot be independently verified. That is not a flaw in this context.\n\nWho is this for? Electroweak and top physicists, and anyone tracking lepton-flavour-universality tests. It is a solid, mature measurement that deserves a serious referee. I would accept it with comments asking for more detail on the fake-transfer validation, ideally a table showing how R_tau/e shifts under alternative transfer assumptions.\n\nRecommendation: send to peer review. This is exactly the kind of measurement that should be in the literature.","headline":"First ATLAS R_tau/e from top-quark decays, consistent with the SM at 2.4% precision; a solid, careful measurement whose only real soft spot is the simulation-only transfer of fake-electron corrections from same-sign to opposite-sign events.","tokens_in":59625,"tokens_out":2367,"would_cite":true,"duration_ms":23687,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["13.38.Be","14.60.Fg"],"model":"deepseek-v4-flash","headline":"New measurement finds tau and electron couplings to the W boson equal within about 2.4% total uncertainty.","keywords":["lepton flavour universality","W boson branching fraction","tau lepton","impact parameter","top quark decay","template fit","proton-proton collisions"],"falsifier":"A measurement of R_tau/e using an independent opposite-sign control region enriched in fake electrons, or a direct data-driven determination of the fake-electron transfer factor, that shifts the central value beyond the quoted systematic uncertainty would challenge the result; likewise, a future more precise measurement that excludes R_tau/e = 1 at the few-percent level would contradict the Standard Model expectation of lepton flavour universality.","tokens_in":58447,"feed_emoji":"⚛️","tokens_out":7180,"duration_ms":60272,"temperature":0.7,"pith_summary":"The paper establishes a measurement of the ratio of W-boson branching fractions to tau leptons and to electrons, R_tau/e = 0.975 ± 0.012 (stat) ± 0.020 (syst). This is consistent with the Standard Model assumption that the W boson couples to all charged leptons with the same strength, a property called lepton flavour universality. The result is the first measurement of this ratio using W bosons produced in top-quark decays, and it improves the precision of the worldwide average for R_tau/e. If correct, it confirms that any difference between tau and electron couplings to the W boson is smaller than about 2.4%, tightening the experimental constraints on new-physics models that violate lepton flavour universality.","feed_headline":"Tau and electron W-couplings equal within 2.4 percent","feed_subtitle":"First measurement of R_tau/e from top-quark decays matches the Standard Model and tightens the world average.","key_machinery":"The central object is the ratio R_tau/e of W-boson branching fractions. The analysis uses a tag-and-probe selection: one lepton tags a top-quark decay in ttbar events and the other electron probes the W decay. The two sources, direct W -> e nu and W -> tau nu -> e nu nu nu, are separated by a two-dimensional template fit to the electron transverse momentum and the absolute value of the transverse impact parameter, the distance of closest approach of the electron track to the beamline. Electrons from tau decays have lower momenta because neutrinos carry energy, and larger impact parameters because the tau lives briefly before decaying; the prompt-electron templates are calibrated with a data sample of Z -> e+ e- events, while tau-decay templates come from simulation corrected for the measured impact-parameter resolution. The ratio is extracted as the parameter of interest in a profile likelihood with more than 300 nuisance parameters.","core_discovery":"The central claim, stated as equation (5) of the paper, is R_tau/e = 0.975 ± 0.012 ± 0.020, with a total uncertainty of 0.024. The measurement uses top-quark pair events where one W boson decays to a tau (which then decays to an electron and neutrinos) and the other W decays directly to an electron; the two cases are separated by the electron's transverse momentum and its impact parameter with respect to the beamline. The value agrees with the Standard Model prediction R_tau/e = 1 and with the previous proton-proton collider measurement, while differing from the combined electron-positron collider result by more than two standard deviations. The paper argues that the measurement will reduce the uncertainty in the world average for R_tau/e.","pith_inferences":["A natural next step is to combine R_tau/e with R_mu/e to test the full triangle of lepton-flavour ratios; if all three agree with unity at sub-percent level, new physics in W-lepton couplings would be constrained below current sensitivity.","The same impact-parameter template method could be applied to measure R_tau/mu directly using muons from tau decays, avoiding electron identification uncertainties entirely.","If the discrepancy with the electron-positron collider result persists as more data accumulate, it may point to a genuine effect or to a systematic in one of the collider environments; the added collision data will adjudicate.","The quoted uncertainty includes the rate of tau -> e nu nu; a more precise measurement of this branching fraction would directly improve R_tau/e in this channel."],"forward_implications":["If correct, the tau and electron couplings to the W boson are equal within about 2.4% total uncertainty.","The measurement contributes to the world average for R_tau/e, improving its precision.","It agrees with the previous proton-proton collider measurement and disagrees with the electron-positron collider combination by more than two standard deviations, so the tension in the world data remains.","Combined with the already precise R_mu/e measurement, it sharpens the overall test of lepton flavour universality in W-boson decays.","The use of the electron impact parameter distribution as the discriminating variable is demonstrated as a practical technique for this channel."],"supporting_citations":[{"why":"Provides the combined electron-positron collider result for R_tau/e that this measurement differs from by over two standard deviations.","marker":"[9]"},{"why":"Provides the previous proton-proton collider measurement of R_tau/e that this result agrees with.","marker":"[11]"},{"why":"The companion measurement of R_mu/e in W-boson decays that this analysis complements.","marker":"[12]"},{"why":"The earlier tau/mu universality test whose impact-parameter calibration method is adapted here.","marker":"[10]"},{"why":"Supplies the world average, the Z and tau branching fractions, and the tau -> e branching fraction used in the analysis.","marker":"[5]"},{"why":"Provides the precisely known tau lifetime used to correct the simulated impact-parameter templates for tau decays.","marker":"[88]"}],"fun_headline_variants":["ATLAS: W decays to tau and electron agree with Standard Model","W boson treats tau and electron alike, new ATLAS measurement","Top-quark decays yield R_tau/e = 0.975 for W bosons","First W-boson tau/electron ratio from top quarks: 0.975"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that fake-electron correction factors measured in a same-sign control sample can be transported to the opposite-sign signal region, with the transfer factor taken from Monte Carlo simulation rather than an independent data control region.","fun_headline_variants_meta":{"raw":{"variants":["ATLAS: W decays to tau and electron agree with Standard Model","W boson treats tau and electron alike, new ATLAS measurement","Top-quark decays yield R_tau/e = 0.975 for W bosons","First W-boson tau/electron ratio from top quarks: 0.975"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00033,"raw_usage":{"total_tokens":1839,"prompt_tokens":945,"completion_tokens":894,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":561,"completion_tokens_details":{"reasoning_tokens":809}},"tokens_in":561,"tokens_out":894,"duration_ms":9223,"temperature":1.0,"reasoning_tokens":809,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:24:26.214183+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A measurement of R_tau/e using an independent opposite-sign control region enriched in fake electrons, or a direct data-driven determination of the fake-electron transfer factor, that shifts the central value beyond the quoted systematic uncertainty would challenge the result; likewise, a future more precise measurement that excludes R_tau/e = 1 at the few-percent level would contradict the Standard Model expectation of lepton flavour universality.","supporting_citations":[],"review_version":1}