{"id":"dc1912a6-4309-4cc7-b16d-75a80633f42b","arxiv_id":"2501.06070","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A single CMS analysis simultaneously measures the cross sections of tZq, tWZ, and ttZ production at 13 TeV, finding values consistent with the Standard Model within uncertainties.","lead":"CMS reports the first combined measurement of top quarks produced with a Z boson, both singly and in pairs, from CERN's Run 2 data. The results mostly match the Standard Model and may sharpen future searches for new physics.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Reported σ(tZq)=0.81 pb in Eq. (1) conflicts with the paper's own tZq SM prediction (94.2±3.1 fb) and best-fit ratio (0.99±0.13); the consistent value is ~0.081 pb. The headline numerical result cannot be used as printed.","rationale":"The reader identified the nonprompt-lepton transfer factor and DNN signal separation as the weakest assumptions. Those are legitimate concerns for a proceedings that omits the full validation, but they are not the first obstacle an informed reader encounters: Eq. (1) quotes a tZq cross section that is almost an order of magnitude above the same paper's SM prediction and best-fit ratio. This is a direct, checkable inconsistency in the headline result, and it takes precedence because no background systematic can bridge the gap. I keep the conditional verdict rather than rejecting: the underlying physics may well be correct, and a missing leading zero in a proceedings is plausible, but the publication condition must explicitly include correcting or confirming the tZq value against the companion paper. If that numerical check passes, the reader's nonprompt and DNN concerns become the relevant remaining risks. My disagreement with the reader is therefore about which concern is most load-bearing, not about whether the paper needs additional supporting evidence.","tokens_in":4821,"tokens_out":10301,"duration_ms":99257,"concrete_test":"Verify the tZq cross section in Eq. (1) against the paper's own inputs and the companion preprint arXiv:2410.23475: recompute 0.99×94.2 fb ≈ 93.3 fb = 0.0933 pb and compare with the value reported in the companion paper. If the companion paper reports 0.081±0.010 pb (81±10 fb), then Eq. (1) and the Conclusion contain a typographical decimal error that must be corrected before publication; if the companion paper reports 0.81 pb, then the prediction, ratio, and measured value are mutually inconsistent and the fit must be re-derived.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"In Section 3, the paper quotes the tZq SM prediction as 94.2±3.1 fb (0.0942 pb) and reports a best-fit cross-section ratio of 0.99±0.13. Multiplying these gives about 93 fb, i.e. 0.093 pb, yet Eq. (1) reports σ(tZq)=0.81±0.07±0.06 pb and the Conclusion repeats 0.81±0.10 pb. The printed value is 810 fb, roughly 8.7 times larger. The ttZ+tWZ line is internally consistent (SM sum of 0.84 pb and 0.136 pb, ratio 1.17, measured 1.14 pb), so the inconsistency is specific to the tZq headline number. No background or systematic uncertainty can reconcile a factor of about 8.7 with the quoted best-fit ratio. Either the proceedings contains a missing-decimal typo (0.081 pb) or the central numerical result is wrong; in both cases the paper as submitted cannot serve as a reliable statement of the claimed simultaneous measurement.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings paper describes a CMS Run 2 measurement at 13 TeV with 138 fb^-1 of a simultaneous extraction of single-top (tZq) and pair-production (ttZ+tWZ) cross sections in final states with three leptons plus a Z candidate. Events are classified with a multiclass DNN into tZq, ttZ+tWZ, and background categories; inclusive cross sections are obtained from a two-parameter profile likelihood scan, and normalized differential cross sections are presented as functions of five observables. The paper reports sigma(ttZ+tWZ) = 1.14 +/- 0.07 pb and sigma(tZq) = 0.81 +/- 0.10 pb and claims good agreement with the SM except for a low-pT(W) trend in the ttZ+tWZ differential distribution.","tokens_in":5060,"tokens_out":4321,"duration_ms":40739,"significance":"If correct, this would be the first simultaneous measurement of tZq and ttZ+tWZ at the LHC and would provide a useful cross-check of separate CMS results, with a reduced dependence on signal modeling assumptions through the simultaneous fit. The proceedings is concise and directs the reader to the companion paper for technical details, which is appropriate for this format. The paper also quotes SM predictions and prior CMS measurements explicitly. However, as written, the central tZq number is not internally consistent with the quoted SM prediction and measured cross-section ratio, so the result cannot currently be used without a correction.","major_comments":[{"comment":"The quoted tZq cross section is inconsistent with the paper's own inputs. Section 3 gives the SM tZq prediction as 94.2 +/- 3.1 fb and reports the best-fit cross-section ratio as 0.99 +/- 0.13; these imply a measured cross section of about 0.093 pb, yet Eq. (1) and the Conclusion quote sigma(tZq) = 0.81 +/- 0.07 (stat) +/- 0.06 (syst) pb and 0.81 +/- 0.10 pb. The printed value is larger by a factor of about 8.7, and no systematic uncertainty can absorb this discrepancy. The manuscript must either correct a missing-decimal typo (0.081 pb) or re-evaluate the result; as printed, the central numerical claim is not usable.","section":"Section 3, Eq. (1), and Section 4"},{"comment":"The nonprompt-lepton background is a major background in the three-lepton final state, and its estimate relies on a transfer-factor method. The text describes the method only in a qualitative way and reports no transfer-factor uncertainty, closure test, or validation numbers; the validation-region sentence is also internally contradictory (\"outside the Z resonance region\" followed by |m(ll)-m(Z)| < 20 GeV). Since both measured cross sections shift with this background, the proceedings should either quantify the method or state explicitly that these details and the associated uncertainties are provided only in the companion paper [1].","section":"Section 2"},{"comment":"The claim that the simultaneous measurement is \"less dependent on the signal modeling assumptions\" is not supported by the material shown. The DNN is trained on the same simulated signal and background models that are used in the fit, and the paper does not report the DNN input variables, training/validation performance, or a comparison of the fitted signal yields with and without the multiclass classification. Without such checks, the reduced model dependence remains an assertion rather than a demonstrated property.","section":"Section 2 (DNN) and Section 3"}],"minor_comments":[{"comment":"The SM prediction text reads \"840 +/- 100 pb\" for ttZ; this should presumably be \"840 +/- 100 fb\", otherwise the sum with tWZ cannot be consistent with the measured 1.14 pb.","section":"Section 3"},{"comment":"The sentence \"events outside the Z boson resonance region are selected, |m(ll)-m(Z)| < 20 GeV\" is self-contradictory; the intended selection is likely a sideband with |m(ll)-m(Z)| > 20 GeV.","section":"Section 2"},{"comment":"The CMS Preliminary label in Figure 2 reads \"1138 fb\" while the text states 138 fb^-1; this is a typo and should be corrected.","section":"Figure 2"},{"comment":"The two-dimensional likelihood plot has no contour levels, color scale, or axis labels visible in the text; this makes the 2D scan difficult to interpret.","section":"Figure 1"},{"comment":"The reference formatting is inconsistent, particularly for reference [3], and DOI/URL formatting should be harmonized with the journal style.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"To the editor: this is a short proceedings contribution that depends heavily on the companion CMS paper (arXiv:2410.23475). The internal inconsistency in the tZq cross section is the kind of error that should be caught before publication; I recommend asking the authors to confirm whether 0.81 pb is a typo for 0.081 pb. If the companion paper actually reports 0.081 pb, this contribution should be corrected accordingly. The lack of quantitative background validation is less concerning in a proceedings context if the companion paper contains it, but the numeric error must be fixed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [colleague],\n\nI read the CMS proceedings on the first simultaneous tZq and ttZ+tWZ measurement. The main thing to know: the analysis is a real first, but the tZq cross section in Eq. (1) doesn't match the paper's own fit. The fit finds a ratio to the SM of 0.99 ± 0.13 for tZq, and the quoted SM prediction is 94.2 fb. That gives about 0.093 pb. Eq. (1) says 0.81 pb, roughly 8.7 times larger. Likely a missing decimal, but as printed the headline number is wrong. The ttZ SM prediction is also printed as '840 ± 100 pb' when it should be fb.\n\nWhat the paper does well: it gives a concise description of a CMS Run-2 analysis at 138 fb^-1 that extracts both tZq and ttZ+tWZ cross sections from one fit, using a multiclass DNN to separate signals and backgrounds. That simultaneous extraction is a genuine first relative to the separate CMS measurements [4,5,6]. The differential results in five observables are included, and the data-MC agreement is reported candidly, including the low-pT(W) trend for ttZ. The writing is clear and the references, including the companion preprint (2410.23475), are appropriate.\n\nSoft spots: the numerical error is load-bearing and affects the central result; it should have been caught before submission. The proceedings also omits the systematic breakdown, DNN validation, and transfer-factor derivation, so the 'less dependent on signal modeling' claim is not demonstrated here. Those omissions are normal for a proceedings, but the decimal error is not. The nonprompt background transfer factor, a standard method, is described too briefly to judge whether it holds in the signal phase space; for that you need the preprint.\n\nWho this is for: readers who want a conference-summary version of the CMS result and are willing to follow the preprint for details. I would cite the preprint, not this proceedings, for the physics. The proceedings itself needs a correction before it can serve as a reliable record. If it came to me for review, I would ask for the corrected tZq number (and the ttZ unit fix) and then accept it as a proceedings summary. The underlying simultaneous measurement is worth taking seriously; just don't quote Eq. (1) as printed.","headline":"The simultaneous tZq/ttZ+tWZ extraction is a genuine first, but the tZq cross section in Eq. (1) contradicts the paper's own fit by about a factor of 8.7; use the companion preprint and correct the proceedings.","tokens_in":5568,"tokens_out":4943,"would_cite":false,"duration_ms":44575,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The first simultaneous measurement of single and pair top-quark production with a Z boson reports sigma(ttZ+tWZ) = 1.14 +/- 0.07 pb and sigma(tZq) = 0.81 +/- 0.10 pb.","keywords":["top quark","Z boson","tZq","ttZ","tWZ","inclusive and differential cross sections","multiclass neural network","LHC Run 2"],"falsifier":"In the sideband used to validate the nonprompt estimate, $|m(\\ell\\ell)-m(Z)| < 20$ GeV, compare the predicted nonprompt yield from the transfer-factor method with the observed event count; a disagreement beyond the quoted uncertainties would mean the background model is biased and the central cross sections are not reliable. Re-running the simultaneous fit with an independent nonprompt estimator, such as a matrix method, and checking whether the two cross sections move by more than the uncertainties would settle the point.","tokens_in":4627,"feed_emoji":"⚛️","tokens_out":8028,"duration_ms":70576,"temperature":0.7,"pith_summary":"This paper presents the first measurement that extracts single-top and top-pair production in association with a $Z$ boson from the same data with one simultaneous fit. Using 138 fb$^{-1}$ of 13 TeV proton-proton collisions, it measures $\\sigma(t\\bar{t}Z+tWZ) = 1.14 \\pm 0.07$ pb and $\\sigma(tZq) = 0.81 \\pm 0.10$ pb, corresponding to ratios to the standard model predictions of $1.17 \\pm 0.07$ and $0.99 \\pm 0.13$. Normalized differential cross sections are reported for five observables, with good overall agreement to simulation except for a low-$p_T(W)$ trend in the $t\\bar{t}Z+tWZ$ channel. The point of the simultaneous approach is that it is less dependent on signal modeling assumptions and can constrain new physics that changes top-$Z$ couplings across different production modes.","feed_headline":"Top+Z cross sections measured together for first time at LHC","feed_subtitle":"CMS combines ttZ, tWZ and tZq in one fit, cutting model dependence and sharpening new-physics tests.","key_machinery":"A multiclass deep neural network (DNN) is the central classifier: it takes 26 kinematic and multiplicity variables and assigns each event to one of three output nodes, $t\\bar{t}Z+tWZ$, $tZq$, and background, with events placed in the category of highest score. The output distributions are used in a profile likelihood fit with two parameters of interest, one for $\\sigma(t\\bar{t}Z+tWZ)$ and one for $\\sigma(tZq)$; for differential cross sections the same two-parameter fit is performed bin by bin and unfolded with response matrices. Nonprompt leptons, a major background, are estimated with a transfer-factor method in which a misidentification rate measured in a background-enriched region is applied to the signal region. The two-dimensional likelihood scan over the two cross-section ratios to the standard model yields the inclusive results.","core_discovery":"The central result is that $t\\bar{t}Z+tWZ$ and $tZq$ production can be measured together in the three-lepton final state and separated with a multiclass deep neural network. The inclusive cross sections are $\\sigma(t\\bar{t}Z+tWZ) = 1.14 \\pm 0.05\\,(\\text{stat}) \\pm 0.04\\,(\\text{syst})$ pb and $\\sigma(tZq) = 0.81 \\pm 0.07\\,(\\text{stat}) \\pm 0.06\\,(\\text{syst})$ pb within the $70 < m_{\\ell^+\\ell^-} < 110$ GeV window. The differential cross sections, unfolded to parton level, agree with standard model predictions within uncertainties, with one exception: the $t\\bar{t}Z+tWZ$ distribution as a function of $p_T(\\ell_W)$ shows a discrepancy at low $p_T(\\ell_W)$ that the authors note is reminiscent of a similar trend in inclusive $t\\bar{t}$ production. The paper presents this as the first consistent simultaneous constraint on single and pair top+Z production, which sharpens the test of top-$Z$ couplings.","pith_inferences":["The simultaneous-fit strategy could extend naturally to other associated top-quark productions, such as $tH$ or four-top production, wherever two signal processes share a final state and would otherwise be measured in separate analyses.","If the low-$p_T(\\ell_W)$ trend persists in a future larger dataset, it would strengthen the case that higher-order QCD corrections to $t\\bar{t}$+jets also improve $t\\bar{t}Z$ modeling; the paper does not claim this.","A direct closure test of the transfer-factor method in a high-multiplicity validation region, comparing predicted and observed nonprompt yields, would quantify how much of the central values rest on that assumption."],"forward_implications":["A single simultaneous fit now constrains $t\\bar{t}Z$, $tWZ$, and $tZq$ cross sections, reducing the model dependence that separate measurements carry.","The measured inclusive cross sections are consistent with standard model predictions, so these channels do not, by themselves, point to new physics in top-$Z$ couplings.","The normalized differential cross sections for five observables provide reference spectra that future standard model and new-physics calculations can be checked against.","The low-$p_T(\\ell_W)$ discrepancy in $t\\bar{t}Z+tWZ$ mirrors a known trend in inclusive $t\\bar{t}$ production, suggesting the origin may be shared QCD modeling of top-pair production rather than a new-physics effect."],"supporting_citations":[{"why":"The preprint version of this measurement; defines the phase space, DNN categories, and simultaneous fit whose results are summarized here.","marker":"[1]"},{"why":"Supplies the lepton misidentification transfer-factor method and the tZq reference cross section used in the fit.","marker":"[4]"},{"why":"Provides the predicted ttZ cross section used as the SM reference for the ttZ+tWZ parameter.","marker":"[5]"},{"why":"Provides the predicted tWZ cross section, which is combined with [5] for the ttZ+tWZ signal template.","marker":"[6]"}],"fun_headline_variants":["First joint measurement of single and pair top-Z production at LHC","CMS measures top-Z single and pair in one fit for first time","Top-Z processes measured simultaneously for the first time","Neural net joins top-Z single and pair cross sections","Top+Z single and pair production measured together for first time"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The measurement assumes the lepton misidentification rate measured in a background-enriched region transfers unchanged to the high-multiplicity signal region; if that transfer fails, both the $t\\bar{t}Z+tWZ$ and $tZq$ cross sections shift away from their true values.","fun_headline_variants_meta":{"raw":{"variants":["First joint measurement of single and pair top-Z production at LHC","CMS measures top-Z single and pair in one fit for first time","Top-Z processes measured simultaneously for the first time","Neural net joins top-Z single and pair cross sections","Top+Z single and pair production measured together for first time"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000696,"raw_usage":{"total_tokens":3117,"prompt_tokens":886,"completion_tokens":2231,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":502,"completion_tokens_details":{"reasoning_tokens":2147}},"tokens_in":502,"tokens_out":2231,"duration_ms":17911,"temperature":1.0,"reasoning_tokens":2147,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:05:55.540723+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In the sideband used to validate the nonprompt estimate, $|m(\\ell\\ell)-m(Z)| < 20$ GeV, compare the predicted nonprompt yield from the transfer-factor method with the observed event count; a disagreement beyond the quoted uncertainties would mean the background model is biased and the central cross sections are not reliable. Re-running the simultaneous fit with an independent nonprompt estimator, such as a matrix method, and checking whether the two cross sections move by more than the uncertainties would settle the point.","supporting_citations":[],"review_version":1}