{"id":"b79be16f-399d-4ae9-b45c-579bb46ce6a8","arxiv_id":"1908.09100","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A simulated LHC analysis shows that the Matrix Element Method could extract the top-Higgs CP-mixing angle to about one degree with 300 inverse femtobarns, and a 5-sigma discovery would need about 20 inverse femtobarns if detector effects are ignored.","lead":"This paper tests whether a statistical method called the Matrix Element Method can measure the CP-mixing angle of the top quark's Higgs coupling using single top quark plus Higgs production at the LHC. Under idealized detector assumptions, it finds that roughly 20 inverse femtobarns of data would reveal a 5-sigma deviation from the Standard Model.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 20 fb^-1 discovery claim rests on a signal-only, perfect-detector closure test in which the pseudo-data and the MEM likelihood come from the same NLO calculation; adding backgrounds and detector smearing is the untested step that would decide whether the projection is meaningful.","rationale":"The reader's verdict is CONDITIONAL, and my stress-test identifies the same core vulnerability: the sensitivity projections assume signal-only pseudo-data and perfect detection, with no background model. I agree that this is the load-bearing assumption. I would add one layer: the pseudo-data and the likelihood weights are produced by the same NLO calculation, so the exercise is a closure test that cannot expose common errors in the event weights or in the modeling of the final state. That strengthens the need for an independent pseudo-data check, but it does not change the verdict: the paper should either include a background model and detector effects, or clearly label the results as an idealized proof of principle. The NLO cross sections are checked against aMC@NLO and the MEM framework is clearly presented, so the concern is about external validity rather than internal inconsistency. Given the reader already conditioned on this missing baseline, no verdict adjustment is needed.","tokens_in":8898,"tokens_out":6041,"duration_ms":72506,"concrete_test":"Generate pseudo-data not from the same NLO code but from an independent sample: aMC@NLO tH events at alpha=22.5 deg matched to a parton shower and passed through a fast detector simulation with realistic jet-energy smearing, then add the dominant backgrounds (ttbar, W+jets, single-top) normalized to their measured cross sections. Run the same MEM likelihood with NLO signal weights and a realistic transfer function on this mixed sample at L=20, 80, and 300 fb^-1; compare the extracted alpha and the significance relative to alpha=0 with Table I and the 20 fb^-1 5-sigma claim. If the required luminosity moves upward by a large factor, or the 300 fb^-1 significance with 3% signal efficiency falls below the stated 3 sigma, the headline sensitivity should be presented only as an idealized closure bound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claims in Sec. IV depend entirely on the simulated measurement described there: unweighted NLO tH events are generated for alpha=22.5 deg and then analyzed with MEM likelihood weights built from the same NLO calculation, with a delta-function transfer function W(y,x)=delta(y-x) (Eq. 8). There is no background term in the probability density P(x|alpha) (Eq. 5) and no background contribution in the pseudo-data. Consequently, the quoted 5-sigma discovery luminosity of about 20 fb^-1, the precisions in Eq. (10), and Table I measure only how well the MEM can separate two signal hypotheses in a pure, perfectly measured tH sample. They are not LHC projections: in a mixed sample, Eq. (5) is no longer the correct probability density, background events contribute an alpha-independent term that dilutes the signal-vs-SM separation, and the required luminosity can shift substantially. The scale-variation systematic labeled [sys.] in Eq. (10) is a valid perturbative check but does not cover background normalization, PDF uncertainties, or detector response. Because the pseudo-data and the likelihood are generated by the same NLO calculation, the exercise is a closure test; it would not catch a common error in the event-weight definition. The paper's own caveat about signal detection efficiencies shows that detector effects are recognized, but the absence of any background model is the unaddressed part of the argument.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a Standard Model extension in which the top-quark Yukawa coupling to the Higgs boson is a mixture of CP-even and CP-odd terms parameterized by an angle alpha, focusing on pp -> tH production at the LHC. The authors compute fiducial and total cross sections at LO and NLO QCD, show that the inclusive cross section is insensitive to small alpha (up to about 25 degrees), and then apply the Matrix Element Method with NLO event weights to extract alpha. Using pseudo-data generated at NLO for alpha = 22.5 degrees and a perfect-detector transfer function, they report alpha = 22.5 +/- 0.9 degrees (stat.) +0.8/-0.6 degrees (sys.) at 300 fb^-1, and estimate that about 20 fb^-1 would allow a 5-sigma discovery if signal detection efficiencies are ignored, while a 3% efficiency would still allow 3-sigma at 300 fb^-1 and a 5-sigma discovery in the high-luminosity phase.","tokens_in":9186,"tokens_out":5520,"duration_ms":58617,"significance":"If the sensitivity projections were realistic, this would be a valuable demonstration that NLO-accurate Matrix Element Methods can substantially improve the extraction of CP-violating top-Higgs couplings in a very challenging final state. The strengths of the paper include the NLO cross-section calculation, which is cross-checked against aMC@NLO, the transparent parametrization of the fiducial cross section, and the use of a standard maximum-likelihood framework. However, the quantitative claims in Section IV rest on signal-only pseudo-data, a delta-function detector transfer function, and the same NLO calculation used for both event generation and MEM weights. The paper is therefore best read as a closure test of the method rather than as a direct LHC sensitivity projection; the numerical projections are conditional on several idealizations that are not modeled.","major_comments":[{"comment":"The probability density in Eq. (5) contains only the tH signal contribution, and the pseudo-data of Section IV are generated as pure NLO tH events with no background processes. Since Eq. (5) is used as the likelihood in Eq. (7), the quoted discovery luminosity of about 20 fb^-1 and the 3-sigma estimate at 300 fb^-1 in Section IV are sensitivities of the MEM in a pure, perfectly reconstructed signal sample, not LHC projections. In a real measurement the likelihood would need a background component, otherwise the estimator is biased and the required luminosity can change substantially; please either include a background model in both the likelihood and the pseudo-data or explicitly relabel the claims as idealized sensitivity limits.","section":"Section IV, Eq. (5)"},{"comment":"The pseudo-data are generated with the same NLO calculation used to compute the MEM weights, including the same renormalization/factorization scale choice and the same coupling parametrization. This makes the exercise a closure test: a common error in the event-weight definition or in the fixed-order calculation would not be revealed by the likelihood fit, and the quoted statistical uncertainties reflect only the internal consistency of the method. I would like to see at least one validation with an independent event sample, for example aMC@NLO with a different scale choice or a parton-shower-matched sample, before the central sensitivity numbers are used as projections.","section":"Section IV, pseudo-data"},{"comment":"The transfer function W(y,x)=delta(y-x) assumes a perfect detector. The manuscript itself notes in Section III that jet-energy variables should include nontrivial jet-energy scales, yet E_j and E_H enter the likelihood in Eq. (9). The effect of detector resolution on the extracted alpha is therefore not assessed; with realistic energy smearing the statistical precision and the discovery luminosity would change, likely in the direction of degraded sensitivity. Please add a smearing model or move this limitation to the abstract and conclusion.","section":"Section III, Eq. (8)"},{"comment":"The uncertainty labeled [sys.] in Eq. (10) and Table I is obtained only from renormalization/factorization scale variation. This is a valid perturbative check, but it is not a complete theory-systematic uncertainty: PDF uncertainties, background normalization, and detector-related systematics are not included. Calling this quantity 'systematic' without qualification is misleading; either relabel it as a scale uncertainty or extend the systematic budget.","section":"Section IV, Eq. (10) and Table I"}],"minor_comments":[{"comment":"In the sentence 'a determination of the mixing angle could envisaged', 'envisaged' should be 'be envisaged'.","section":"Section IV"},{"comment":"The definition of the top-tagged jet and how the top-quark four-vector is reconstructed are not given, although eta_t is one of the MEM variables in Eq. (9). Please specify the jet algorithm and the reconstruction procedure used to obtain the top-tagged jet.","section":"Section III, Eq. (9)"},{"comment":"The coefficient of cos(alpha)sin(alpha) is quoted as 0.00; the numerical uncertainty on this coefficient should be given so the reader can judge the statement that it is 'compatible with zero'.","section":"Eq. (3)"},{"comment":"The '5-sigma discovery' estimate is stated without specifying the test statistic or how the significance is computed; a short description, such as a likelihood-ratio or Neyman-Pearson construction, would improve reproducibility.","section":"Section IV"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a technically competent study, and the NLO MEM machinery is interesting. The main issue is not the correctness of the NLO calculation but the mismatch between the idealized simulation and the LHC projections advertised in the abstract: the central numbers in Section IV are closure-test sensitivities with no backgrounds, a perfect detector, and the same calculation used for both generation and analysis. I would be willing to accept a revised version that either adds backgrounds/detector effects and a proper systematic budget, or reframes all sensitivity claims as an idealized proof-of-principle with the limitations stated prominently in the abstract and conclusions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid proof-of-principle paper. The genuinely new piece is applying the authors' NLO matrix element weights to tH production with a CP-violating top-Higgs coupling and showing that the MEM can recover an input CP-mixing angle where the inclusive cross section cannot. The NLO cross sections are cross-checked against aMC@NLO, the likelihood setup is standard, and the writing is clear. Credit where due: they don't oversell the cross-section measurement, they give a clean derivation of the alpha dependence, and they correctly note that scale uncertainties hide angles up to about 25 degrees. The MEM result, alpha-hat = 22.5 ± 0.9 stat +0.8/-0.6 sys at 300 fb^-1, is a useful benchmark.\n\nThe soft spot is exactly what the stress-test note says. The pseudo-data are NLO tH signal events, the likelihood has no background term, and the transfer function is delta(y-x) even for jet energies. The quoted 20 fb^-1 for a 5 sigma discovery is therefore a best-case, signal-only number, not an LHC projection. Since pseudo-data and weights come from the same NLO calculation, the exercise is a closure test and would not catch an error common to both. The scale-variation systematic is a valid perturbative check but covers neither background normalization nor detector response. I don't think this is fatal, because the paper is transparent about the perfect-detector assumption and calls the efficiency treatment a first step in spirit. But the abstract's 'about 20 fb^-1' is easy to over-read, and the mismatch between that headline and the idealized setup should be made unavoidable.\n\nMy take: if the paper is read as a methodological proof of principle, it works. If the authors want the discovery numbers to be treated as physics projections, they need at least a background model and a smeared transfer function. For a journal, I would send it to peer review and request that qualification, not reject. It deserves a serious referee. Audience: phenomenological hep-ph readers, people working on MEM or top-Higgs CP. I'd bring it to a reading group as an example of an honest closure test.","headline":"A clean NLO MEM closure test for tH CP mixing; the quoted luminosities are idealized signal-only numbers and should be labeled as such.","tokens_in":9753,"tokens_out":4304,"would_cite":true,"duration_ms":45250,"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 Matrix Element Method at NLO QCD can measure the top-Higgs CP-mixing angle to within about a degree, with about 20 fb⁻¹ of LHC data sufficient for a 5σ deviation from the Standard Model when detection efficiencies are ignored.","keywords":["top-Higgs coupling","CP violation","Matrix Element Method","next-to-leading order QCD","single top production","Higgs Yukawa coupling","LHC phenomenology","CP-mixing angle"],"falsifier":"Generate pseudo-data that include the dominant Standard Model backgrounds (e.g., $t\\bar t H$, $tWH$, $t\\bar t$+jets, $W$+jets) and a realistic detector transfer function with jet-energy smearing, then rerun the NLO Matrix Element likelihood extraction; if the recovered $\\alpha$ is biased by more than the quoted $\\pm 0.9^\\circ$ at $300~\\mathrm{fb}^{-1}$, or if the luminosity needed for a $5\\sigma$ signal rises well above $20~\\mathrm{fb}^{-1}$, the paper's idealized projections would fail under real LHC conditions.","tokens_in":8644,"feed_emoji":"⚛️","tokens_out":20668,"duration_ms":165680,"temperature":0.7,"pith_summary":"This paper asks whether the LHC can directly reveal the nature of the Higgs boson's interaction with the top quark — in particular, whether that interaction is the pure Standard Model coupling or a mixture of the two parity possibilities, parametrized by a mixing angle $\\alpha$. The answer it argues is yes, provided single-top-plus-Higgs ($pp\\to tH$) events are analyzed with the Matrix Element Method at next-to-leading order in QCD. A plain cross-section measurement cannot tell the Standard Model from mixing angles below about $25^\\circ$, but the method recovers a benchmark input of $\\alpha = 22.5^\\circ$ with a $\\pm0.9^\\circ$ statistical uncertainty at $300~\\mathrm{fb}^{-1}$, and roughly $20~\\mathrm{fb}^{-1}$ of data would establish a $5\\sigma$ deviation from the Standard Model under idealized detector and signal-only assumptions. This matters because modified top-Higgs couplings are among the most plausible early signs of new physics, and this process is one of the few direct handles on them.","feed_headline":"20 fb^-1 would reveal a CP-violating top-Higgs coupling","feed_subtitle":"Sub-degree precision on the top-Higgs CP angle at 300 fb^-1, where plain cross-section counting fails.","key_machinery":"The load-bearing mechanism is the Matrix Element Method likelihood: for each event with kinematic variables $x$, a weight $P(x|\\alpha) = \\frac{1}{\\sigma(\\alpha)}\\int d^n y\\, \\frac{d^n\\sigma(y|\\alpha)}{dy_1\\cdots dy_n}\\, W(y,x)$ is computed from fixed-order NLO QCD matrix elements, and the estimator $\\hat\\alpha$ is read off from the minimum of the negative log-likelihood over the event sample. The paper assumes a perfect detector, $W(y,x)=\\delta(y-x)$, and describes each event by seven kinematic variables: the pseudo-rapidity of the top-tagged jet, and the energy, pseudo-rapidity, and azimuthal angle of the hardest light jet and of the Higgs boson. All weights are generated at NLO QCD accuracy because, as the paper stresses, the NLO corrections to $pp\\to tH$ are large (up to about $+32\\%$) and depend on $\\alpha$, so a leading-order implementation would be unreliable. A secondary but essential piece is the CP-violating Yukawa parametrization with $a=1$, $b=2/3$, which keeps the $gg\\to H$ cross section fixed and lets $\\alpha$ interpolate continuously between the CP-even ($\\alpha=0^\\circ$) and CP-odd ($\\alpha=180^\\circ$) limits.","core_discovery":"The central claim is that the Matrix Element Method, evaluated with fixed-order NLO QCD event weights, turns the rare $pp\\to tH$ process into a precision probe of the top-Higgs interaction. The paper works with a specific beyond-Standard-Model scenario in which the top-quark Yukawa coupling is a mixture of CP-even and CP-odd terms, $L_{t\\bar t H} = -\\frac{y_t}{\\sqrt{2}}(a\\cos\\alpha\\,\\bar t t + i b\\sin\\alpha\\,\\bar t\\gamma_5 t)H$, with $a=1$ and $b=2/3$ chosen so that the gluon-fusion Higgs cross section is unchanged for any $\\alpha$. In this scenario the inclusive cross section cannot separate the Standard Model from $\\alpha$ values below about $25^\\circ$, because the interference terms are suppressed by the fiducial cuts, but the likelihood built from NLO event weights recovers a benchmark input of $\\alpha = 22.5^\\circ$ as $\\hat\\alpha = 22.5^\\circ \\pm 0.9^\\circ$ (stat.) $^{+0.8^\\circ}_{-0.6^\\circ}$ (sys.) at $300~\\mathrm{fb}^{-1}$. The authors further estimate that about $20~\\mathrm{fb}^{-1}$ would give a $5\\sigma$ signal in the idealized case of a perfect detector and signal-only pseudo-data, while a few-percent signal efficiency would push discovery to the high-luminosity LHC, with theory uncertainties no longer limiting beyond roughly $L \\approx 425~\\mathrm{fb}^{-1}$.","pith_inferences":["Replacing the perfect-detector transfer function $\\delta(y-x)$ with realistic jet-energy smearing would likely widen the statistical uncertainty on $\\hat\\alpha$; the quoted sub-degree precision should be read as the method's idealized resolving power in the best case.","With no backgrounds in the pseudo-data, the $20~\\mathrm{fb}^{-1}$ discovery figure is a lower bound; including $t\\bar t H$, $tWH$, and multijet backgrounds could plausibly raise the required luminosity by an order of magnitude, an effect the paper does not quantify.","The near-zero $\\sin\\alpha$ and $\\cos\\alpha\\sin\\alpha$ coefficients in the fiducial cross section are a consequence of the chosen phase-space cuts; a different cut design that preserves the CP-odd phase-space region could make even the inclusive cross section sensitive to $\\alpha$, an option the paper leaves implicit.","The per-event likelihood could be combined with the measured $t\\bar t H$ cross section in a joint fit to separate the size and sign of the top-Higgs coupling, disentangling degeneracies that neither measurement alone can resolve; the paper does not perform this combination."],"forward_implications":["The CP-mixing angle $\\alpha$ becomes a measurable LHC observable: a sub-degree determination from $pp\\to tH$ events is possible with $300~\\mathrm{fb}^{-1}$, and the extraction stops improving with statistics beyond roughly $L \\approx 425~\\mathrm{fb}^{-1}$, where the constant $\\pm 0.7^\\circ$ scale-uncertainty term dominates.","A $5\\sigma$ discovery of a CP-violating top-Higgs coupling at the benchmark level $\\alpha = 22.5^\\circ$ requires only about $20~\\mathrm{fb}^{-1}$ in the idealized signal-only case; with a few-percent signal efficiency, $300~\\mathrm{fb}^{-1}$ gives $3\\sigma$ and the high-luminosity LHC reaches $5\\sigma$.","Because the likelihood weights are fixed-order NLO QCD predictions, the method inherits their accuracy; the paper shows this matters, since the NLO corrections are large (up to about $+32\\%$) and $\\alpha$-dependent, so leading-order analyses of this process would be unreliable.","The same machinery can be applied to other rare processes and to alternative BSM scenarios such as two-Higgs-doublet models, which the paper notes as an equivalent target for the technique."],"supporting_citations":[{"why":"Supplies the CP-violating top-Higgs coupling parametrization with mixing angle α that defines the BSM scenario.","marker":"[28]"},{"why":"Provides the a = 1, b = 2/3 parameter choice that keeps the gluon-fusion Higgs cross section fixed, plus the earlier tH phenomenology this analysis extends.","marker":"[21]"},{"why":"Establishes the extension of the Matrix Element Method beyond the Born approximation, the event-weight calculus the likelihoods rest on.","marker":"[36]"},{"why":"Applies the NLO Matrix Element Method to single top-quark production at the LHC, the technical template for this study.","marker":"[37]"},{"why":"Provides NLO event-weight predictions for jet events defined by 2→1 jet algorithms, used for the fixed-order weights here.","marker":"[39]"},{"why":"The automated NLO framework whose differential results cross-check the authors' calculation.","marker":"[30]"},{"why":"The 13 TeV tH search whose few-percent signal-efficiency estimate drives the realistic luminosity projections.","marker":"[27]"},{"why":"Introduces the dynamical likelihood (Matrix Element) method that the whole analysis applies to tH production.","marker":"[31, 32]"}],"fun_headline_variants":["MEM achieves sub-degree CP angle in top-Higgs at 300 fb^-1","20 fb^-1 enough for CP-violating top-Higgs signal","Cross sections blind to CP; Matrix Element Method sees it","Precision probe of top-Higgs CP from rare tH events","Matrix Element Method turns rare tH into CP precision tool"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the paper's own idealization, stated at Eq. (8) and in the Section IV pseudo-data setup: a perfect detector and signal-only NLO events with no background, so all quoted luminosities and the sub-degree precision apply only to a signal-only, perfect-detector world.","fun_headline_variants_meta":{"raw":{"variants":["MEM achieves sub-degree CP angle in top-Higgs at 300 fb^-1","20 fb^-1 enough for CP-violating top-Higgs signal","Cross sections blind to CP; Matrix Element Method sees it","Precision probe of top-Higgs CP from rare tH events","Matrix Element Method turns rare tH into CP precision tool"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000306,"raw_usage":{"total_tokens":1792,"prompt_tokens":1025,"completion_tokens":767,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":641,"completion_tokens_details":{"reasoning_tokens":674}},"tokens_in":641,"tokens_out":767,"duration_ms":7864,"temperature":1.0,"reasoning_tokens":674,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:21:15.651813+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Generate pseudo-data that include the dominant Standard Model backgrounds (e.g., $t\\bar t H$, $tWH$, $t\\bar t$+jets, $W$+jets) and a realistic detector transfer function with jet-energy smearing, then rerun the NLO Matrix Element likelihood extraction; if the recovered $\\alpha$ is biased by more than the quoted $\\pm 0.9^\\circ$ at $300~\\mathrm{fb}^{-1}$, or if the luminosity needed for a $5\\sigma$ signal rises well above $20~\\mathrm{fb}^{-1}$, the paper's idealized projections would fail under real LHC conditions.","supporting_citations":[{"cited_title":"A framework for Higgs characteri- sation,","cited_arxiv_id":null,"evidence_quote":"Supplies the CP-violating top-Higgs coupling parametrization with mixing angle α that defines the BSM scenario."},{"cited_title":"Higgs production in association with a single top quark at the LHC,","cited_arxiv_id":null,"evidence_quote":"Provides the a = 1, b = 2/3 parameter choice that keeps the gluon-fusion Higgs cross section fixed, plus the earlier tH phenomenology this analysis extends."},{"cited_title":"Extending the Matrix Ele- ment Method beyond the Born approximation: Calcu- lating event weights at next-to-leading order accuracy,","cited_arxiv_id":null,"evidence_quote":"Establishes the extension of the Matrix Element Method beyond the Born approximation, the event-weight calculus the likelihoods rest on."},{"cited_title":"The Matrix Element Method at next-to-leading order QCD for hadronic collisions: Single top-quark production at the LHC as an example appli- cation,","cited_arxiv_id":null,"evidence_quote":"Applies the NLO Matrix Element Method to single top-quark production at the LHC, the technical template for this study."},{"cited_title":"The automated computation of tree- level and next-to-leading order differential cross sections, and their matching to parton shower simulations,","cited_arxiv_id":null,"evidence_quote":"The automated NLO framework whose differential results cross-check the authors' calculation."},{"cited_title":"Search for associated production of a Higgs boson and a single top quark in proton-proton collisions at√s = 13 TeV,","cited_arxiv_id":null,"evidence_quote":"The 13 TeV tH search whose few-percent signal-efficiency estimate drives the realistic luminosity projections."}],"review_version":1}