{"id":"2edf54f8-b1d5-4b90-9293-3afbafbf2e58","arxiv_id":"2412.04357","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"low","formal_verification":"none","parameter_count":2,"one_line_summary":"Using 138 fb^-1 of 13 TeV proton-proton collisions, CMS finds no tau-pair resonance and sets its strongest limit yet on Z' bosons.","lead":"CMS looked for a new heavy particle, the Z' boson, decaying into two tau leptons using the full LHC Run 2 dataset. No excess over standard model predictions was seen, and the search excludes a Z' below 3.5 TeV in a common benchmark model.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified.","rationale":"The reader's weakest assumption points to the right part of the analysis: the QCD ABCD estimate is the most data-driven and least directly validated component. However, the paper itself provides a quantitative argument that this component cannot affect the headline exclusion: QCD is a small fraction of the background in the high-m_rec tail that determines the 3.5 TeV bound, and the very large shape uncertainties assigned there are shown to have small impact. The central claim therefore does not hinge on the linearity of the QCD transfer factors. The remaining background estimates (DY, W+jets, tt, VV, single top) are simulation-based with data-derived scale factors, and the control-region plots show reasonable agreement. No internal inconsistency, circular step, or missing support that would invalidate the exclusion was found. A direct robustness check with an alternative QCD transfer-factor model is still worth running, but nothing in the current text requires changing the reader's ACCEPT verdict.","tokens_in":41049,"tokens_out":9528,"duration_ms":118761,"concrete_test":"Using the public HEPData tables, recompute the combined 95% CL sigma x B limits with the QCD transfer-factor shape uncertainty replaced by a quadratic or by per-bin nonparametric transfer factors. If the observed limit crosses the SSM prediction at an m_Z' shifted by more than about 100 GeV from 3.5 TeV, the QCD-shape concern would land; otherwise it is confirmed to be negligible.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The closest candidate for a load-bearing concern is the ABCD estimate of the QCD multijet background, which the reader also identifies. Section 7.1 states that the required lack of correlation between the charge-sign and tau-identification observables is validated only in a sideband with pTmiss < 30 GeV, and Section 8 applies closure tests only in that low-pTmiss sideband. The transfer factors are fitted with first-order polynomials in m_rec, so a nonlinear dependence could, in principle, bias the QCD shape in the signal region. This concern is explicitly weighed by the paper: the QCD contribution is 6–9% of the tau_l tau_h signal-region yields and 27–35% of the tau_h tau_h yield, but the latter is concentrated at low m_rec and is of low relevance to the high-mass signal scenarios. The shape systematic uncertainties reach 150% at high m_rec, yet the paper argues that the small QCD fraction there makes the impact on the limits small. Since the central claim, the exclusion of an SSM Z' below 3.5 TeV, is governed by the high-m_rec tail where non-QCD backgrounds dominate and are validated by data control regions, this concern does not threaten the central result.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports a search for narrow heavy neutral resonances (Z') decaying to tau lepton pairs using 138 fb^-1 of proton-proton collision data at sqrt(s)=13 TeV recorded by CMS. Three final states are analyzed: tau_mu tau_h, tau_e tau_h, and tau_h tau_h. The discriminating variable is an approximated reconstructed mass m_rec(Z') built from the visible tau decay products and the inferred neutrino transverse momentum. Backgrounds are estimated from simulation normalized with data control regions, with a data-driven ABCD method for QCD multijet production. A binned maximum-likelihood fit to the m_rec(Z') distributions finds the observed yields consistent with the standard-model expectation. Upper limits at 95% CL are set on sigma B(Z' -> tau tau) as a function of m_Z', and in the sequential standard model scenario a Z' with mass below 3.5 TeV is excluded, the most stringent limit to date from this type of search.","tokens_in":41281,"tokens_out":7775,"duration_ms":84835,"significance":"If the result holds, it provides the strongest direct constraint on heavy neutral resonances decaying to tau lepton pairs, extending previous ditau bounds by roughly 1 TeV. The analysis has clear strengths: three decay channels are combined with per-year scale factors, the systematic uncertainties are documented and correlated appropriately, the limit-setting procedure uses the standard CLs method implemented in the CMS combine tool, and tabulated results are provided in HEPData. The background model is data-driven for the dominant QCD component in the tau_h tau_h channel and simulation-based with control-region scale factors for the electroweak backgrounds. The principal methodological caveat, the extrapolation of ABCD transfer factors from low to high pTmiss and from low to high m_rec, is explicitly quantified and is shown to have only small impact on the central exclusion because the QCD multijet fraction is small in the high-m_rec region that governs the 3.5 TeV bound.","major_comments":[],"minor_comments":[{"comment":"The QCD ABCD method is validated only in a low-pTmiss sideband (pTmiss < 30 GeV), and the lack-of-correlation check is documented for the hadronic di-tau channel, while the signal-region transfer factors are used for the lepton+tau channels and for pTmiss > 30 GeV; a sentence explaining why the low-pTmiss validation is sufficient, or a cross-check in a high-pTmiss sideband, would increase confidence in the QCD background model.","section":"Section 7.1 and Section 8"},{"comment":"The symbols pZ'miss and p tau_vis_i are used both as three-vectors and as magnitudes in the same expression; adding explicit vector notation and defining E_tau_vis_i as the visible energy of each tau daughter would remove ambiguity.","section":"Section 2, Eq. (1)"},{"comment":"The sentence stating that the theoretical decay width increases with g_tau but remains smaller than the experimental resolution (<30%) should specify that this is the width relative to the Z' mass, since Section 2 quotes a <3% relative width for the SSM scenario; otherwise the two statements appear inconsistent.","section":"Section 5"},{"comment":"The description of the QCD transfer-factor shape uncertainty would be clearer if it stated explicitly that the uncertainty is taken as the full deviation of the fitted first-order polynomial from unity, rather than the statistical uncertainty of the fit, given that the quoted shape uncertainties (up to 150%) are much larger than the transfer-factor statistical uncertainties quoted in the same section.","section":"Section 8"}],"recommendation":"accept","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The bottom line is simple: this is a well-executed, standard LHC search with a null result. What is actually new is the full Run-2 dataset, the three-channel combination (mu-tau, e-tau, tau-tau), and the resulting SSM Z' mass exclusion of 3.5 TeV, which improves on the previous 2.4 TeV bound. The reconstructed-mass estimator and background methods are reused from earlier CMS tau-pair analyses, but that is not a defect; the analysis is a straightforward extension with more data, done carefully.\n\nThe paper's strengths are the things that make LHC searches trustworthy: per-channel and per-year scale factors derived from dedicated control regions, detailed systematic accounting, a transparent ML fit with nuisance parameters, and HEPData tables for the limits. The QCD multijet estimate uses an ABCD method with transfer factors fitted as first-order polynomials in m_rec. The validation and closure tests are performed in a low-pTmiss sideband, and the shape uncertainties reach 150% at the high end of the m_rec distribution. That sounds concerning, but the stress-test note is right: in the high-mass tail where the 3.5 TeV exclusion lives, QCD is a small fraction (6-9% in the mixed channels, and the tau-tau QCD is concentrated at low mass). The central result is driven by DY and W+jets backgrounds that are validated with data control regions. So the QCD extrapolation is a minor soft spot, not a load-bearing one.\n\nI would have liked to see a more direct discussion of the linearity assumption for the transfer factors, but the shape systematic partially covers this, and the small QCD fraction makes the impact negligible for the stated limits. There is no circularity: the limit comes from comparing data to a background-only prediction, and the signal templates are simulated at fixed masses.\n\nWho should read this? Phenomenologists working on third-generation-motivated Z' models, and anyone putting together tau-pair resonance exclusions. It is not a breakthrough, but it is the current best bound in this channel and it is honestly reported. A serious referee would find no fatal flaws; the paper already passed peer review at Physical Review D. I would accept it.","headline":"A solid, incremental Run-2 search: no signal, and the tau-pair Z' exclusion improves from about 2.4 to 3.5 TeV in the SSM, based on a careful three-channel combination of the full 138 fb^-1 dataset.","tokens_in":41767,"tokens_out":1409,"would_cite":true,"duration_ms":18020,"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":"This paper reports a search for heavy neutral $Z'$ bosons decaying to tau lepton pairs in 138 fb$^{-1}$ of 13 TeV proton-proton collisions, finds the data consistent with standard model predictions, and excludes a sequential standard…","keywords":["Z' boson search","tau lepton pair final state","heavy neutral gauge boson","sequential standard model","CMS detector","LHC Run 2","95% confidence level exclusion","reconstructed mass estimator"],"falsifier":"Measure the QCD transfer factor $N_C/N_D$ in finely binned $m_{\\mathrm{rec}}(Z')$ intervals using a high-statistics validation region with low missing transverse momentum; if the measured transfer factor deviates from the first-order polynomial by more than the quoted shape uncertainty in the signal region, the background prediction and the derived limit would need to be revised.","tokens_in":40873,"feed_emoji":"⚛️","tokens_out":9731,"duration_ms":87075,"temperature":0.7,"pith_summary":"This paper reports a search for a heavy neutral gauge boson, the $Z'$, decaying to tau lepton pairs, using 138 fb$^{-1}$ of proton-proton collisions at 13 TeV recorded by the CMS detector. The observed reconstructed-mass distributions in the three tau-decay channels agree with standard model background predictions, so the paper claims no new resonance. It sets upper limits at 95% confidence level on the production cross section times branching fraction for $Z'$ masses from 250 GeV to 5 TeV. In the sequential standard model benchmark, a $Z'$ with mass below 3.5 TeV is excluded, and the paper states this is the most stringent limit to date for a $Z'$ decaying to tau pairs. If correct, the result closes the tau-pair window on $Z'$ bosons with standard-model-like couplings below 3.5 TeV and constrains models with enhanced third-generation couplings.","feed_headline":"Tau-pair search excludes Z' bosons below 3.5 TeV","feed_subtitle":"The most stringent bound yet on a heavy neutral boson decaying to tau leptons.","key_machinery":"The discriminating variable is the reconstructed $Z'$ mass estimator $m_{\\mathrm{rec}}(Z') = \\sqrt{ (E^{\\tau\\mathrm{vis}}_1 + E^{\\tau\\mathrm{vis}}_2 + |\\mathbf{p}^{Z'\\mathrm{miss}}_T|)^2 - |\\mathbf{p}^{\\tau\\mathrm{vis}}_1 + \\mathbf{p}^{\\tau\\mathrm{vis}}_2 + \\mathbf{p}^{Z'\\mathrm{miss}}|^2 }$, built from the visible products of the two tau decays plus an assumed neutrino momentum that balances the visible transverse momentum. This estimator is used as the fit discriminant in a profile maximum-likelihood fit to extract the signal yield. Standard model backgrounds come from simulation rescaled by data-measured scale factors, while the QCD multijet background is estimated with an ABCD method whose transfer factors are fitted as first-order polynomials in $m_{\\mathrm{rec}}(Z')$ to assign the shape uncertainty.","core_discovery":"The central result is a null observation: after a maximum-likelihood fit to the reconstructed mass estimator $m_{\\mathrm{rec}}(Z')$ in the $\\tau_\\mu\\tau_h$, $\\tau_e\\tau_h$, and $\\tau_h\\tau_h$ channels, the data are consistent with standard model expectation. The resulting 95% confidence level upper limits on $\\sigma(pp\\to Z')\\,\\mathcal{B}(Z'\\to\\tau^+\\tau^-)$ exclude a sequential standard model $Z'$ with mass below 3.5 TeV; for tau-pair branching fractions of 1% and 10%, the excluded masses are 3.0 TeV and 4.1 TeV, respectively. The $\\tau_h\\tau_h$ channel dominates the sensitivity over most of the mass range, while the mixed $\\tau_\\mu\\tau_h$ and $\\tau_e\\tau_h$ channels become competitive at $Z'$ masses of 4–5 TeV because the hadronic tau identification efficiency degrades for highly boosted decays.","pith_inferences":["Indirectly, the null result tightens constraints on models with nonuniversal couplings to third-generation fermions, since those predict an enhanced tau-pair rate relative to the standard model $Z$.","A testable extension is to reinterpret the HEPData tables in models with $Z$-$Z'$ interference or with mass-dependent couplings; the paper notes interference effects are small for the benchmark range.","With the higher luminosity of the high-luminosity LHC, the same reconstructed-mass estimator and background strategy should push the tau-pair exclusion toward higher masses, assuming the QCD transfer-factor systematics do not grow with reach.","The estimator's reliance on transverse-momentum balance could be reused in searches for similar resonances in final states with only one visible tau and missing energy, where the neutrino momentum is even less constrained."],"forward_implications":["Any sequential standard model $Z'$ boson with $Z$-like couplings and mass below 3.5 TeV is excluded at 95% confidence level, improving the previous tau-pair bound.","The published limits on $\\sigma\\mathcal{B}(Z'\\to\\tau^+\\tau^-)$ as a function of $m_{Z'}$ let any model prediction be tested against the data directly.","Models predicting a tau-pair branching fraction of 1% are excluded below 3.0 TeV, and those predicting 10% are excluded below 4.1 TeV.","The $\\tau_h\\tau_h$ channel gives the dominant sensitivity across most of the mass range, while at 4–5 TeV the mixed channels take over because of tau identification efficiency at high boost.","No channel shows a significant excess, so the observations are consistent with the standard model background prediction in every mass bin."],"supporting_citations":[{"why":"Prior CMS tau-pair search whose 2.42 TeV bound this analysis explicitly improves.","marker":"[9]"},{"why":"CMS dilepton search that sets the 5.15 TeV benchmark exclusion for an SSM Z' and provides the comparison context.","marker":"[1]"},{"why":"ATLAS dilepton search with the 5.10 TeV exclusion that the new tau-pair limit is compared against.","marker":"[5]"},{"why":"Defines the simplified phenomenological model with free Z' mass and couplings used to generate the signal samples.","marker":"[25]"},{"why":"Provides the hadrons-plus-strips algorithm used to reconstruct hadronic tau decays.","marker":"[41]"},{"why":"Supplies the DeepTau deep neural network that separates genuine hadronic taus from jets, electrons, and muons.","marker":"[42]"}],"fun_headline_variants":["No sign of Z' bosons in tau-pair searches at 13 TeV","Z' bosons below 3.5 TeV excluded in tau-pair final states","Tau-pair search sets strongest limit on heavy Z' bosons","No Z' signal: CMS excludes heavy resonances in tau pairs","Heavy neutral boson to tau pairs: ruled out below 3.5 TeV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the QCD multijet background in the signal region can be extrapolated from an ABCD sideband using transfer factors that are fitted as first-order polynomials in the reconstructed mass; if those transfer factors curve nonlinearly with mass, the background shape and the derived 95% confidence level limit would be biased.","fun_headline_variants_meta":{"raw":{"variants":["No sign of Z' bosons in tau-pair searches at 13 TeV","Z' bosons below 3.5 TeV excluded in tau-pair final states","Tau-pair search sets strongest limit on heavy Z' bosons","No Z' signal: CMS excludes heavy resonances in tau pairs","Heavy neutral boson to tau pairs: ruled out below 3.5 TeV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000677,"raw_usage":{"total_tokens":3056,"prompt_tokens":898,"completion_tokens":2158,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":514,"completion_tokens_details":{"reasoning_tokens":2062}},"tokens_in":514,"tokens_out":2158,"duration_ms":14268,"temperature":1.0,"reasoning_tokens":2062,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T21:30:19.038118+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the QCD transfer factor $N_C/N_D$ in finely binned $m_{\\mathrm{rec}}(Z')$ intervals using a high-statistics validation region with low missing transverse momentum; if the measured transfer factor deviates from the first-order polynomial by more than the quoted shape uncertainty in the signal region, the background prediction and the derived limit would need to be revised.","supporting_citations":[],"review_version":1}