{"id":"133d89f1-93fa-4b5f-803a-a2201cc4f8cb","arxiv_id":"hep-ph/9905386","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Next-to-leading-order QCD predictions for W and Z boson pair production with correlated leptonic decays are presented for the Tevatron at 2 TeV and the LHC at 14 TeV, validated against two earlier independent calculations.","lead":"The authors compute, including one-loop QCD corrections and full decay correlations, how often hadron colliders produce W and Z boson pairs that decay to leptons. These are the dominant backgrounds to Higgs and supersymmetry searches, so the numbers set how sensitive those searches can be.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The width-insertion prescription of §3 multiplies the whole amplitude, including singly-resonant and interference terms, and no numerical error estimate is given for Tables 5–7; the effect looks bounded at the few-percent level, so the headline numbers probably survive.","rationale":"The reader's weakest_assumption points at exactly the right place: the §3 width prescription and the absence of any numerical estimate of its effect on Tables 5–7. I agree that this is the softest joint in the paper, and I agree with the reader's overall disposition, because the concern is bounded rather than structural. Three independent considerations support keeping ACCEPT. First, the external benchmark is genuine: Tables 1 and 2 compare eight Born and eight full entries against Frixione–Nason–Mele–Ridolfi at √s = 16 and 40 TeV and agree at the 0.1–1% level, which is an independent-group cross-check of the doubly-resonant machinery, the parton distributions, the αs running of Eq. 1 and the narrow-width division by branching ratios. Second, the numbers are internally consistent: folding the Table 3 total of 13.0 pb for W⁺W⁻ at 2 TeV through two leptonic branching ratios (~1/9 each) gives ~160 fb before acceptance, and the 70.9 fb of Table 5 corresponds to a ~45% acceptance under the pT > 20 GeV, |η| < 2, MET > 25 GeV cuts of Eqs. 4–5, which is plausible; the same arithmetic at the LHC (120.6 pb → ~1480 fb → 514 fb) is likewise consistent. Third, the width prescription is not an invention of this paper — it is the published Baur–Vermaseren–Zeppenfeld treatment cited as [18], so under the citation rule the scheme is supported; what is missing is a sensitivity study, not a justification. Where I would push slightly beyond the reader is on the internal tension: §4.1 and Fig. 4 argue that the singly-resonant Wγ* contribution raises the trilepton rate from 42.6 to 49.4 fb, i.e. that this class of diagrams matters at the ~16% level, while §3 argues the mistreatment of that same class is confined to a region where it does not matter. Both statements can be true simultaneously — the γ*-dominated region is far off the Z pole, where the multiplicative factor is close to unity — but the paper never reconciles them quantitatively, and that reconciliation is precisely what the concrete test would supply. A secondary and weaker point, not load-bearing for the headline: the §4.1 conclusion that 'studies using PYTHIA therefore underestimate the standard model background' rests on an internal Z-only versus Z/γ* comparison within an NLO calculation, not on a like-for-like comparison with the LO PYTHIA numbers of [20], so the attribution of the discrepancy to the missing Wγ* alone is not fully established. That affects a phenomenological aside, not Tables 3–6. Neither issue moves the verdict; both would be addressed by a short numerical study of the width scheme, which is the single check most worth running before these numbers are used as the standard reference background. The paper's other acknowledged limitations — the omitted τ decays in the Table 7 comparison with Han–Turcot–Zhang, the LO treatment of t t̄, the deferred details of the subtraction implementation in §1 — are stated openly and are appropriately scoped by the authors, so I do not treat them as un","tokens_in":13871,"tokens_out":4299,"duration_ms":75716,"concrete_test":"Recompute two entries with the width factor applied only to the doubly-resonant amplitude (equivalently, a complex-mass scheme for the resonant propagators), leaving the singly-resonant and interference terms with the plain 1/(q²−M²) propagators: the WZ→e⁻e⁺ν entry of Table 5 (2.89 fb, MRS98) and the WW→e⁻ν̄e⁺ν entry (70.9 fb). If either shifts by more than about 3%, the three-significant-figure precision quoted in Tables 5–6 is not supported and the numbers experimental searches subtract need a stated systematic; if both shift by less than 1%, the §3 approximation is quantitatively harmless and the concern does not land.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's own §3 states that the prescription — multiplying the entire amplitude by the product over propagators of (q²−M²)/(q²−M²+iMΓ) — 'is the correct treatment for the doubly-resonant piece, but mistreats the singly-resonant diagrams.' Every entry in Tables 5, 6 and 7 contains singly-resonant diagrams, and no number anywhere quantifies the size of the mistreatment. The technical sharpness of the concern is that the factor is applied to the full amplitude, so the DR–SR interference term is multiplied by a product of propagator factors that has no diagrammatic justification; near q²≈M² that factor carries a phase and a magnitude as small as 1/√2, so the interference term is rotated and damped in exactly the region where it is largest. The authors' defence — that the mistreatment is confined to q²≈M² where the doubly-resonant piece dominates — is structurally reasonable, and the low-dilepton-mass region that drives the Wγ* enhancement of §4.1 and Fig. 4 is far off the Z pole, where the factor tends to unity. So the concern is real but bounded, not fatal. Two things keep it from being dismissed outright: (i) the paper elsewhere argues the singly-resonant γ* piece is numerically significant (49.4 fb versus 42.6 fb, a ~16% effect in the trilepton sample), so the class of diagrams being 'mistreated' is not negligible by the authors' own accounting; and (ii) the quoted precision of Tables 5–6 (three significant figures, e.g. 2.89 fb and 70.9 fb) is finer than the accuracy the scheme can claim. The prescription itself is not novel — it follows Baur, Vermaseren and Zeppenfeld [18], a published source — so the objection is to the absence of a quantitative error estimate, not to the scheme's legitimacy.","agreement_with_reader":"agree"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The paper presents a next-to-leading-order QCD calculation of vector-boson pair production at hadron colliders (W+W-, W±Z/γ*, Z/γ*Z/γ*) with the leptonic decays of the massive bosons included, so that all spin correlations are retained and experimental cuts can be imposed on the four-lepton final state. The calculation is implemented in the Monte Carlo program MCFM, based on the Catani-Seymour subtraction formalism. The authors first validate against the earlier Frixione-Nason-Mele-Ridolfi results at √s = 16 and 40 TeV using the HMRSB distributions (Tables 1 and 2), then give updated total cross sections at √s = 2 TeV and 14 TeV with MRS98 and CTEQ5 (Tables 3 and 4). Section 3 extends the treatment beyond the zero-width approximation by including singly-resonant diagrams with a gauge-preserving width insertion, and Tables 5 and 6 quote NLO cross sections for specific leptonic channels after pT, η and missing-ET cuts. Section 4 gives two applications: the WZ/γ* trilepton background to SUSY gaugino production, and a parton-level di-lepton Higgs search with the WW, tt, ZZ and WZ backgrounds. An appendix collects the tree-level helicity amplitudes for the doubly- and singly-resonant contributions.","tokens_in":14141,"tokens_out":4171,"duration_ms":75868,"significance":"This is a substantial and timely update. The genuine advance is the combination of the full O(αs) corrections with the leptonic decay correlations in a single calculational framework, which is what makes the numbers usable as backgrounds for Run II and LHC searches rather than as inclusive on-shell rates. The external validation in Tables 1 and 2 is unusually strong for a calculation of this kind: eight independent Born and full entries agree with the Frixione-Nason-Mele-Ridolfi numbers at the 0.1–1% level, which checks both the Born and the real/virtual machinery against an independent group. The internal consistency is also good: the WW, ZZ and WZ total cross sections of Table 3 (13.0, 1.56 and 1.95+2.01 pb) reproduce the normalizations quoted in Table 7. The paper is candid about its approximations, and the two worked examples in §4 are exactly the kind of practical guidance the experimental community needs. The results are falsifiable and, in principle, reproducible from the amplitudes given in the appendix.","major_comments":[{"comment":"The prescription multiplies the entire amplitude by the product over propagators of (q²−M²)/(q²−M²+iMΓ), and the authors state plainly that this 'is the correct treatment for the doubly-resonant piece, but mistreats the singly-resonant diagrams.' Every entry in Tables 5, 6 and 7 contains singly-resonant diagrams and therefore inherits this approximation, yet no numerical estimate of its size appears anywhere in the paper. The concern is not academic: the factor is applied to the full amplitude, so the DR–SR interference term is multiplied by a product of propagator factors with no diagrammatic justification, and near q²≈M² that factor carries both a phase and a magnitude as small as 1/√2. The authors' argument that the mistreatment is confined to q²≈M², where the doubly-resonant piece dominates, is structurally reasonable, and the low-mass region driving the Wγ* enhancement of §4.1 lies far from the Z pole where the factor tends to unity. But the paper's own §4.1 shows that the singly-resonant γ* class is a ~16% effect in the trilepton sample (49.4 fb versus 42.6 fb), so the class of diagrams being mistreated is not negligible by the authors' own accounting. I would ask for at least one quantitative handle on the size of the effect — for example, a comparison of Tables 5 and 6 with the width factor applied only to the doubly-resonant piece, or a variation of the prescription — so that the reader can judge whether the three-significant-figure entries are meaningful at that precision.","section":"§3, width insertion prescription","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text states that the cuts of Eqs. (4) and (5) 'suffice for WW production; for the other cases we perform a mass cut 75 < M_{l+l-} < 105 GeV.' For the W±Z/γ* and Z/γ*Z/γ* entries this window removes most of the γ* contribution that §4.1 argues is numerically important, so the WZ and ZZ numbers in Tables 5 and 6 are not directly comparable to the trilepton result of §4.1 and Fig. 4, which use the very different cut set of Eq. (7). The tables should state explicitly which entries are γ*-suppressed by the mass window, and ideally quote the corresponding numbers without it, so that a reader using these as backgrounds knows which photon contribution is and is not included.","section":"§3.1, Tables 5 and 6","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The sentence 'We see that as the invariant mass of deviates from the Z-peak' is missing a word; presumably 'of the lepton pair'.","section":"§4.1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phrase 'an partial set of diagrams' should read 'a partial set of diagrams'.","section":"§3, paragraph after the width discussion","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The entry 'ZZ → e−e+(νν × 3)' is ambiguous. Please state whether this sums over the three neutrino flavours, whether the Z→νν branching ratio is included, and whether the missing-ET cut of Eq. (5) is applied to the neutrino pair.","section":"Table 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caveats that the tt background is implemented only at leading order (footnote 2) and that τ decays are not included are important for the comparison with Ref. [24] and should appear in the table caption, not only in the body text.","section":"Table 7 and §4.2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The diagram labels in the version I received are not legible. Please ensure the final figure has readable particle labels, since the doubly- versus singly-resonant distinction is the point of the figure.","section":"Figure 3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"CTEQ5 is cited as MSU-HEP-903100 and hep-ph/9903282. Please update to the published reference if it is now available.","section":"Reference [11]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The statement that the Run II K-factor is 'approximately 1.3 for each case' is loose given the tabulated values (1.30, 1.34, 1.28); a parenthetical range would be more informative.","section":"§2.2","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"This is a solid and important paper that belongs in the journal; the validation against the Frixione-Nason-Mele-Ridolfi results is convincing and the applications are well chosen. My recommendation of minor revision reflects two local additions — a quantitative estimate of the width-prescription uncertainty and clarity about the mass window in Tables 5 and 6 — rather than any doubt about the central result. The paper promises further details of MCFM in a subsequent publication; the editor may wish to encourage that the code be made publicly available, since the numerical tables cannot be independently checked from the text alone."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Bottom line: this is the reference calculation for diboson production at Run II and the LHC, and it earns that status. The new content is real — first complete O(alpha_s) corrections with leptonic spin correlations retained through the decays, which refs [2–7] did not have — and the appendix is not filler. The singly-resonant helicity amplitudes (Eqs. 22, 27, 33, 37, and the B7 functions of Eqs. 28–29) are genuinely new: once a gluon is radiated, the relabelling argument of Dixon-Kunszt-Signer no longer supplies the singly-resonant piece, so those functions had to be built. That is the part of the paper a phenomenologist will actually read line by line, and it looks correct to me on inspection of the spinor structure and the helicity-flip operations in Eqs. 39 and the DR analogues. The validation is the strongest thing here: Tables 1 and 2 compare eight Born and full entries against an independent group at 16 and 40 TeV and agree at the 0.1–1% level. That is an external benchmark, not a self-consistency check, and it is what makes the rest credible. The K-factors of ~1.3 (Run II) and 1.3–1.7 (LHC) are the practical payload. On the soft spots, the stress-test lands but is bounded, and I'd say it slightly overstates itself. Yes, §3 multiplies the whole amplitude by the propagator product, so the DR–SR interference picks up a phase-carrying factor with no diagrammatic justification. But that factor departs from unity only near q²≈M², which is exactly where the doubly-resonant piece dominates; the low-mass region driving the Wγ* enhancement in Fig. 4 (49.4 vs 42.6 fb) sits far off the Z pole, where the factor tends to one. The authors state the limitation plainly and their defence is structurally sound. What I would actually press on is narrower: Tables 5–6 quote three significant figures (2.89 fb, 70.9 fb), which is finer than the scheme can honestly claim. Minor, but worth a sentence in revision. Table 7's comparison with Han-Turcot-Zhang omits tau decays and the t-tbar background is LO only (footnote 2) — both acknowledged, both minor. The real weak axis is reproducibility: §1 defers the method to a later publication and the code was not public at the time. Historically that resolved itself, but a referee should note it. Worth flagging that the abstract does not oversell; the limitations are stated in the body, not buried. Send it out. Anyone working on Higgs-to-WW, trilepton SUSY backgrounds, or diboson phenomenology needs this in their citation list, and the amplitude appendix stands on its own.","headline":"The paper that made diboson backgrounds quantitative — the NLO-plus-decay-correlations numbers in Tables 3–6 are the ones the experiments actually subtracted, and the external validation against Frixione-Nason-Mele-Ridolfi holds up.","tokens_in":14942,"tokens_out":1534,"would_cite":true,"duration_ms":26288,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["12.38.Bx","13.85.-t","13.85.Qk","14.70.Fj","14.70.Hp"],"model":"selfhosted/deepseek-v4.1-flash","headline":"The paper computes the full one-loop QCD corrections to W and Z pair production at hadron colliders, keeping the spin correlations of the decaying bosons so that experimental cuts can be applied to the final leptons.","keywords":["vector boson pair production","next-to-leading-order QCD","W and Z bosons","hadron collider backgrounds","lepton spin correlations","finite-width treatment","trilepton supersymmetry background","Higgs search backgrounds"],"falsifier":"Recompute Tables 5, 6 and 7 with an alternative gauge-invariant width scheme, such as the fermion-loop scheme or the pole scheme, and compare channel by channel: a shift larger than the few-percent scale the paper attributes to parton distribution choice would show that the quoted background rates carry an unquantified systematic. An independent check is the Tevatron Run II measurement of the $W^+W^-$ cross section in the di-lepton plus missing-energy channel, which can be compared directly with the $70.9$ fb prediction after the cuts of Eqs. (4) and (5).","tokens_in":15123,"feed_emoji":"⚛️","tokens_out":3346,"duration_ms":60359,"temperature":0.7,"pith_summary":"Vector-boson pair production is the dominant Standard Model background to a long list of new-physics searches, and at the Tevatron its cross section is small enough that theory error matters as much as experimental error. This paper's aim is to supply the missing ingredient: a next-to-leading-order QCD calculation in which the W and Z bosons are not treated as stable on-shell particles but are decayed to leptons with all spin correlations retained, so that realistic acceptance cuts can be imposed on the observed four-lepton (or two-lepton plus missing-energy) final state. The authors implement this in a Monte Carlo program, cross-check their total rates against the earlier calculations of Frixione, Nason, Mele and Ridolfi, and then publish updated numbers for $pbar p$ at $sqrt{s}=2$ TeV and $pp$ at $sqrt{s}=14$ TeV. If the numbers hold, they are the ones experimental searches must subtract: $13.0$ pb for $W^+W^-$ at Run II rising to $120.6$ pb at the LHC with the MRS98 set, and after lepton cuts $70.9$ fb versus $514$ fb in the $e nu_e e nu$ channel. The paper also argues that the photon-mediated contribution to $W Z/gamma^*$ is not negligible, which changes how large the trilepton background to supersymmetry searches actually is.","feed_headline":"W and Z pairs: full one-loop rates for Run II and the LHC","feed_subtitle":"One-loop QCD corrections with lepton spin correlations raise the Standard Model diboson backgrounds by 30 to 70 percent.","key_machinery":"The load-bearing object is a next-to-leading-order Monte Carlo (MCFM) built on the dipole subtraction formalism of Catani and Seymour, fed by the six- and seven-particle helicity amplitudes of Dixon, Kunszt and Signer for the doubly-resonant diagrams plus a new set of sub-amplitudes $B^{tree,a}_7$ and $B^{tree,b}_7$ for the singly-resonant ones, which are obtained by relabelling the external legs of the same functions. Two further pieces carry weight. The propagator factors $P_{ij} = s_{ij}/(s_{ij}-M_{ij}^2)$ organize which diagrams are resonant, and the finite width is introduced not by modifying each propagator independently but by computing with $1/(q^2-M^2)$ and then multiplying the whole amplitude by $prod (q^2-M^2)/(q^2-M^2+iMGamma)$ over the resonant propagators, a prescription that preserves gauge invariance by construction. That trick is what allows the singly-resonant diagrams to be included at all, and it is also the source of the paper's main internal approximation.","core_discovery":"The central claim is that a full $O(alpha_s)$ calculation of $qbar q to V_1 V_2 to$ four leptons, with helicity amplitudes that keep every spin correlation through the decay, gives diboson cross sections that are meaningfully different from what the older literature and the existing event generators provide. At the Tevatron the $K$-factor is about $1.3$ for each channel and at the LHC it runs from about $1.3$ for $Z$ pairs to $1.7$ for $Z W^-$, so the QCD corrections are not a small perturbation on the Born rates. Two consequences are drawn out. First, the authors go beyond the zero-width approximation by including singly-resonant diagrams alongside the doubly-resonant ones, and they show in the trilepton channel that the off-shell photon contribution dominates the low dilepton-mass region, so analyses that included only the $Z$ (as the event generators then did) underestimate the Standard Model background in exactly the signal window used for supersymmetry. Second, applying the same machinery to a Higgs search in the $H to W^{(*)}W^{(*)}$ channel, they find all diboson backgrounds larger than a comparable PYTHIA-based study, by roughly a factor of two for the $WZ$-class events, while the signal itself is also about 20% larger, so the net effect on $S/sqrt{B}$ may be modest.","pith_inferences":["If the multiplicative width prescription misestimates the singly-resonant diagrams at the several-percent level, that error lands preferentially in the off-shell regions that new-physics searches use as signal windows, so the practical size of the effect is worth more than its formal status as an all-orders ambiguity.","The comparison with PYTHIA-based studies suggests a general lesson beyond this paper: any search background quoted from a parton-shower generator that omits photon-mediated diboson diagrams and interference should be re-derived, since the missing piece is concentrated in the low-mass sidebands where background-suppression cuts often sit.","The fact that the signal and the background both grew by comparable factors relative to the earlier analysis implies that $S/sqrt{B}$ is a blunt instrument for judging whether a background update matters; the shape of the dilepton mass distribution carries more information than the total rate.","Having the full spin correlations in the decay means the same machinery could be re-run with arbitrary polarization or angular observables imposed on the leptons, which is where anomalous triple-gauge-coupling searches would look for deviations from these Standard Model predictions."],"forward_implications":["Experimental searches at Run II and the LHC get a set of NLO diboson background numbers with the same lepton acceptance cuts they apply, replacing leading-order or parton-level estimates that were typically 30% low at the Tevatron.","The $W Z/gamma^*$ channel, not $WZ$ alone, must be counted as background to trilepton supersymmetry searches, because the off-shell photon dominates the low invariant-mass region where the signal was defined.","The residual scale dependence at next-to-leading order is small enough that the spread between modern parton distribution sets, about 3% at Run II and 6% at the LHC, becomes a leading theory uncertainty on these cross sections.","Higgs searches in the $140$-$190$ GeV window have their dominant background fixed by the $W^+W^-$ rate, which the paper puts at $13.0$ pb total and $185$ fb after the di-lepton cuts of the analysis it compares to.","Because the singly-resonant diagrams are included, the calculation can be used to model the high-mass tails of the Breit-Wigner distributions that matter when a new-physics signal sits far off the $Z$ peak."],"supporting_citations":[{"why":"Supplies the helicity amplitudes for the doubly-resonant six- and seven-particle processes, and the loop prescription that the singly-resonant amplitudes are built from by relabelling.","marker":"[1]"},{"why":"Earlier next-to-leading-order $ZZ$ calculation used as the numerical benchmark the new program must reproduce in the on-shell limit.","marker":"[2]"},{"why":"Earlier $WZ$ and $W^+W^-$ calculation providing the comparison values in Tables 1 and 2, with a modified version supplied for the $ZW^pm$ case.","marker":"[3]"},{"why":"Earlier $W^+W^-$ calculation at 16 and 40 TeV that anchors the cross-check of the new total rates.","marker":"[4]"},{"why":"Catani-Seymour dipole subtraction, the method on which the next-to-leading-order Monte Carlo is constructed.","marker":"[8]"},{"why":"The Ellis-Ross-Terrano subtraction technique underlying the phase-space treatment of the real-emission corrections.","marker":"[9]"},{"why":"MRS98 parton distribution set, one of the two used for all the Run II and LHC numbers in Tables 3 through 6.","marker":"[10]"},{"why":"CTEQ5 parton distribution set, the second choice, whose spread against MRS98 sets the quoted 3% to 6% theory band.","marker":"[11]"},{"why":"Source of the gauge-preserving prescription for inserting a finite width by multiplying the amplitude with propagator ratios, which the whole off-shell treatment rests on.","marker":"[18]"},{"why":"Provides the trilepton cuts and the signal window against which the $WZ/gamma^*$ background estimate is compared.","marker":"[20]"}],"fun_headline_variants":["Diboson rates at one loop: K-factors of 1.3 to 1.7","One-loop QCD lifts W and Z pair rates by 30 to 70 percent","Off-shell photon diagrams fill the low-mass trilepton window","Full one-loop diboson rates for Tevatron Run II and the LHC","Diboson backgrounds for Higgs searches, recomputed at one loop"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The finite-width treatment multiplies the entire amplitude, singly-resonant diagrams included, by a product of Breit-Wigner ratios, and the authors state plainly that this is correct only for the doubly-resonant piece and mistreats the singly-resonant diagrams near $q^2 approx M^2$, with no numerical estimate of the size of that error anywhere in the paper.","fun_headline_variants_meta":{"raw":{"variants":["Diboson rates at one loop: K-factors of 1.3 to 1.7","One-loop QCD lifts W and Z pair rates by 30 to 70 percent","Off-shell photon diagrams fill the low-mass trilepton window","Full one-loop diboson rates for Tevatron Run II and the LHC","Diboson backgrounds for Higgs searches, recomputed at one loop"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1167,"prompt_tokens":981,"completion_tokens":186,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":91}},"tokens_in":981,"tokens_out":186,"duration_ms":3132,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":91,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T18:29:33.765285+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Recompute Tables 5, 6 and 7 with an alternative gauge-invariant width scheme, such as the fermion-loop scheme or the pole scheme, and compare channel by channel: a shift larger than the few-percent scale the paper attributes to parton distribution choice would show that the quoted background rates carry an unquantified systematic. An independent check is the Tevatron Run II measurement of the $W^+W^-$ cross section in the di-lepton plus missing-energy channel, which can be compared directly with the $70.9$ fb prediction after the cuts of Eqs. (4) and (5).","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the helicity amplitudes for the doubly-resonant six- and seven-particle processes, and the loop prescription that the singly-resonant amplitudes are built from by relabelling."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier next-to-leading-order $ZZ$ calculation used as the numerical benchmark the new program must reproduce in the on-shell limit."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier $WZ$ and $W^+W^-$ calculation providing the comparison values in Tables 1 and 2, with a modified version supplied for the $ZW^pm$ case."},{"cited_title":"Frixione, Nucl","cited_arxiv_id":null,"evidence_quote":"Earlier $W^+W^-$ calculation at 16 and 40 TeV that anchors the cross-check of the new total rates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Catani-Seymour dipole subtraction, the method on which the next-to-leading-order Monte Carlo is constructed."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The Ellis-Ross-Terrano subtraction technique underlying the phase-space treatment of the real-emission corrections."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"MRS98 parton distribution set, one of the two used for all the Run II and LHC numbers in Tables 3 through 6."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"CTEQ5 parton distribution set, the second choice, whose spread against MRS98 sets the quoted 3% to 6% theory band."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the gauge-preserving prescription for inserting a finite width by multiplying the amplitude with propagator ratios, which the whole off-shell treatment rests on."},{"cited_title":"Supersymmetry Reach of the Tevatron via Trilepton, Like-Sign Dilepton and Dilepton plus Tau Jet Signatures","cited_arxiv_id":"hep-ph/9904282","evidence_quote":"Provides the trilepton cuts and the signal window against which the $WZ/gamma^*$ background estimate is compared."}],"review_version":1}