{"id":"81fc4e62-7195-48aa-9b9a-21c7d5f48d11","arxiv_id":"2412.16685","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Updated QCD prediction for top-pair production near threshold at 13 TeV: toponium bound-state effects plus NLL threshold resummation enhance the mass distribution by about 10% and give 12 pb in the 340 to 350 GeV bin versus 8.4 pb from POWHEG.","lead":"This paper updates the Standard Model prediction for top-quark pairs forming a short-lived toponium state near threshold at the LHC, adding NLL resummation of soft and collinear gluons. The new prediction is about 10% higher than fixed-order QCD in the peak region, which matters for interpreting a CMS excess near twice the top quark mass.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'conservative lower bound' recommendation rests on an unproven transfer of the LP/NLP bounding property from DIS/Higgs to top-pair hadroproduction; a direct NLO-level check can settle this.","rationale":"The reader's weakest-assumption analysis identifies precisely the load-bearing step: the lower-bound property is transferred from DIS and Higgs production to top-pair hadroproduction without proof or numerical verification. My reading confirms that this transfer is essential to the paper's framing of the prediction as a conservative background input. If the LP/NLP bounding property fails for the octet or qqbar channels, or if the Mellin-space and z-space power expansions differ enough to flip the inequality, then the central recommendation is unsupported even though the NLL resummation and Coulomb Green's function computation may be internally correct. I considered whether a more severe concern exists, such as the heuristic matching between NRQCD and fixed-order regions or the absence of soft-scale variation. Those are real limitations, but the paper is transparent about them and they do not by themselves invalidate the central numerical prediction. The lower-bound claim, by contrast, is asserted as a justification for adopting the prediction in experimental analyses, and it rests on an extrapolation that is neither proven nor numerically checked. The proposed test is feasible: the NLO hard functions are already computed, so comparing the full NLO Mellin moments with their LP truncation for the three channels would immediately show whether the bounding pattern holds at the first nontrivial order. I therefore agree with the reader's CONDITIONAL verdict: the paper's calculation is plausible and useful, but the conservative lower-bound framing needs an explicit check before the prediction is adopted as the recommended SM input.","tokens_in":9915,"tokens_out":3477,"duration_ms":34976,"concrete_test":"Directly test the LP/NLP bounding property at NLO for each of the three channels. Compute the exact NLO hard function F_i,j->T in Mellin-N space and compare it with its leading-power large-N approximation and with the LP+NLP truncation over the relevant range of N (roughly N = 1-100 for LHC kinematics). If, for any channel, the LP truncation exceeds the exact NLO result or the LP+NLP truncation falls below it, the lower-bound claim fails already at NLO for this process. As a complementary check, use the existing NNLO fixed-order results from MATRIX (already shown in Fig. 3) to verify whether the NLO+NLL NRQCD prediction lies below the NNLO fixed-order curve in the threshold region; if it does not, the claim that the prediction is a conservative lower bound would be contradicted by the paper's own fixed-order comparison.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's recommendation that the NLO+NLL NRQCD prediction be used as a conservative SM background input depends on the claim, made in the paragraph beginning 'In passing let us also address concerns raised in Ref. [10]', that 'the LP and NLP terms provide a lower and upper bound on the exact result', with support cited only for DIS and Higgs boson production in gluon-gluon fusion (Refs. [23,24]). The transfer to top-quark pair hadroproduction is not established for the three channels actually used: gg -> 1S0[1], gg -> 1S0[8], and qqbar -> 3S1[8]. The color-octet channels involve different color structures and a repulsive Coulomb potential, and the qqbar-initiated channel has no direct analogue in the cited processes. Moreover, footnote 1 concedes that the Mellin-N power expansion used here differs from the z-space expansion of Ref. [10] by 'numerically important sub-leading terms', which undermines any assumption that the sign pattern of LP/NLP terms survives unchanged. This matters because the lower-bound statement is used to label the 12 pb integrated prediction in [340,350] GeV as conservative relative to the 8.4 pb POWHEG background; if the bounding property fails, the prediction may be neither a lower bound nor safely conservative, even if the NLL calculation itself is correct. The paper's additional argument that missing NNLO corrections are expected to enhance the cross section is an estimate, not a proof, and is not equivalent to the LP/NLP bound.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents updated NRQCD-based predictions for the top-quark pair invariant mass distribution near threshold at the LHC (13 TeV). Using the framework of Ref. [8], the authors compute the dominant channels gg -> 1S0[1], gg -> 1S0[8], and q qbar -> 3S1[8], combining Coulomb resummation through non-relativistic Green's functions with NLL threshold resummation of the hard functions in Mellin space. They compare the NLO+NLL results with NLO, fixed-order NNLO, and POWHEG predictions, finding an enhancement in the threshold region (12 pb versus 8.4 pb in the [340,350] GeV bin), and they propose using their results as a conservative SM background input for experimental analyses. They also argue that their prediction constitutes a lower bound on the cross section, based on LP/NLP considerations and on estimates of missing NNLO corrections.","tokens_in":10239,"tokens_out":9700,"duration_ms":82411,"significance":"If the calculation is correct, this is a valuable theory input for CMS and ATLAS searches in the top-pair threshold region, with a concrete numerical difference from the currently used POWHEG-based background. The paper is transparent about its framework and provides an ancillary file with the NLO+NLL predictions, which is a strength. The central quantitative claim is a direct calculation and is not fitted to the CMS excess. However, the 'conservative lower bound' interpretation is not yet established, and the scale-stability statement is not supported by the tables for all channels. The paper would be strengthened by either removing or properly qualifying the lower-bound claim, and by clarifying the channel content of the quoted integrated cross section.","major_comments":[{"comment":"The claim that 'the LP and NLP terms provide a lower and upper bound on the exact result' is supported only by references to DIS and Higgs boson production in gluon-gluon fusion (Refs. [23,24]), and is then applied without proof to top-quark pair hadroproduction. The channels used here include gg -> 1S0[8] and q qbar -> 3S1[8], which involve color-octet final states with a repulsive potential and a q qbar initial state that has no direct analogue in the cited processes. Moreover, footnote 1 concedes that the Mellin N-space power expansion used here differs from the z-space expansion of Ref. [10] by 'numerically important sub-leading terms', so the sign pattern of LP/NLP terms found in Refs. [23,24] cannot simply be assumed to transfer. This matters because the statement 'the results presented here can be considered as a lower bound on the cross section' is used to characterize the 12 pb integrated prediction in [340,350] GeV as conservative relative to the 8.4 pb POWHEG result. Please either remove the lower-bound claim or verify it numerically for each of the three channels, for example by comparing the exact NLO result with LP-only and LP+NLP truncations in this process.","section":"Paragraph beginning 'In passing let us also address concerns raised in Ref. [10]'"},{"comment":"The text states that 'the scale stability of the resummed results is somewhat improved', but Tables I and II show the opposite for the channel that produces the toponium resonance, gg -> 1S0[1]. For M_ttbar = 2 m_t, the NLO values across mu_r = mu_f = m_t, 2 m_t, 4 m_t are 18.2, 18.7, 18.3 (range 0.5) while the resummed values are 19.4, 20.5, 21.1 (range 1.7). For M_ttbar = 2 m_t - 5 GeV, the corresponding ranges are 0.4 and 1.9. Thus the resummation does not improve scale stability in this channel, and may worsen it. Please qualify the claim or explain why this is not a concern for the stated improvement.","section":"Tables I and II and paragraph following Table II"},{"comment":"The sentence 'Both these effects are expected to enhance the cross section, so that our current NRQCD prediction can be considered as an estimate for the lower bound' combines two estimates (NNLO corrections to the Green's function and to the hard functions, each O(10%)) with the LP/NLP bound. These are estimates based on e+e- studies and on expected higher-order behavior, not demonstrated bounds for this process. As written, the 'lower bound' conclusion is a conjecture rather than a proven statement. If the authors intend to keep this phrasing, they should provide a concrete numerical check or clearly label the statement as an expectation rather than a bound.","section":"Final paragraph before 'The impact of different parameter settings'"},{"comment":"The quoted integrated cross section of 12 pb in [340,350] GeV and the curves labeled 'sum' in Figures 1-4 include only the three channels of Table I. The text states that other S-wave channels (gg -> 3S1[1,8], q qbar -> 1S0[1,8], gq -> 1S0[1,8] and gq -> 3S1[8]) contribute an additional O(5%). Since the POWHEG comparison includes all channels, the 12 pb versus 8.4 pb comparison is not fully inclusive. Please either include the missing channels in the plotted sum and in the integrated bin, or explicitly state that the quoted numbers exclude them and estimate the impact on the comparison.","section":"Paragraph beginning 'In a next step, we compute the full partonic and hadronic cross sections' and paragraph with the…"}],"minor_comments":[{"comment":"The inverse Mellin integrand in Eq. (5) uses x^{-N}, whereas the forward transform in Eq. (4) is defined with z^{N-1} and the natural inverse is z^{-N}; please make the variable consistent and define x if it is not z.","section":"Equation (5)"},{"comment":"Typo: 'preocess' should be 'process' in the sentence 'From studies of the related preocess of e+e- annihilation'.","section":"Discussion around e+e- annihilation"},{"comment":"Figure 4 shows a 'recommended transition' between NRQCD and POWHEG, but the text does not provide a concrete formula or algorithm for the additive matching; please state how the transition is implemented (for example, a switching function or a hard cutoff) so that the proposal is reproducible.","section":"Figure 4 and matching proposal"},{"comment":"The statement that the dynamical scale H_T/2 is close to m_t near threshold is plausible but not quantified; a short numerical estimate of H_T/2 in the threshold region would be helpful.","section":"Footnote 2"},{"comment":"Please state explicitly in the ancillary file or its documentation whether the tabulated NLO+NLL predictions include only the three dominant channels, and which scale choices and PDF set are used.","section":"Ancillary file"}],"recommendation":"major_revision","confidential_remarks":"The paper is a useful update of an established framework, but the lower-bound language is stronger than the evidence presented. The novelty relative to Ref. [8] is mostly incremental (updated PDFs, scale choices, and a comparison to the CMS analysis), so the editor may wish to consider whether the letter format is appropriate for the amount of new material. If the lower-bound claim is removed or properly qualified, the paper could become acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a solid computational update with real new numbers, but the packaging goes one step too far. The paper recomputes toponium-enhanced ttbar production in NRQCD with NLL threshold resummation for Run 2 conditions, and the new tables and ancillary files are useful. The comparison against POWHEG near 2mt is exactly what experimentalists need. Credit where due: the authors state their limitations plainly, including missing NNLO corrections and fixed soft scale.\n\nWhere it stumbles: the claim that the result is a 'lower bound' on the cross section. That rests on LP/NLP bounding proved in DIS and gg->H, not on the three channels here (gg singlet, gg octet, qqbar octet). The octet channels have repulsive Coulomb and different color structures, and qqbar has no direct analogue. The paper's footnote even admits the Mellin-space expansion differs from the SCET z-space one by numerically important sub-leading terms, which cuts against assuming the sign pattern survives. So the lower-bound wording is unsupported as stated. This matters because the 12 pb vs 8.4 pb comparison is presented as conservative. The actual NLL calculation may well be right, and the missing NNLO pieces are argued to push up, so the prediction is probably an underestimate. But 'probably' is not a proof.\n\nAlso minor: scale uncertainty not varied for mu_s, matching to POWHEG is heuristic, no code released (though ancillary data files exist). These are acknowledged, so not fatal.\n\nBottom line: for readers working on top-quark thresholds or BSM ttbar searches, this is a useful, honest update. The central numbers are plausible and the paper deserves referee time. The authors should either verify the LP/NLP bound for these channels at NLO or weaken the claim. I'd send to peer review.","headline":"Solid computational update with useful new numbers, but the 'lower bound' label is overclaimed and should be either verified or softened before the paper is used as an experimental input.","tokens_in":10785,"tokens_out":1399,"would_cite":true,"duration_ms":13785,"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":"Standard Model QCD, including toponium bound-state effects, predicts a top-pair background near 345 GeV of 12 pb, about 40% above the current Monte Carlo estimate.","keywords":["top-quark pair production","toponium","threshold resummation","Coulomb resummation","NRQCD","invariant mass distribution","LHC","heavy Higgs search"],"falsifier":"The transfer of the LP/NLP bounding property can be tested directly: compute the exact NLO (or NNLO) threshold corrections including next-to-leading-power terms for the three channels $gg\\to {}^1S_0^{[1]}$, $gg\\to {}^1S_0^{[8]}$, and $q\\bar q\\to {}^3S_1^{[8]}$ in Mellin space and check whether the NLO+NLL resummed result lies below them in every channel. A single channel where it lies above would falsify the lower-bound claim; alternatively, a measurement of the 340–350 GeV bin at the LHC with total uncertainty below about 20% would discriminate 12 pb from 8.4 pb directly.","tokens_in":9719,"feed_emoji":"⚛️","tokens_out":16370,"duration_ms":116500,"temperature":0.7,"pith_summary":"This paper updates the Standard Model prediction for top-quark pair production near the threshold where the pair's invariant mass is about twice the top-quark mass, matching the Run 2 kinematics and settings of the LHC search for heavy Higgs bosons decaying to top pairs. It includes Coulomb resummation for the quasi-bound 'toponium' state formed by gluon exchange between top and anti-top, combined with threshold resummation of soft and collinear gluon emissions at next-to-leading-logarithmic accuracy. The central result is that the Standard Model background in the 340–350 GeV bin is predicted to be about 12 pb, compared with about 8.4 pb from the Monte Carlo generator predictions used by the experimental analysis. If correct, this means the excess seen near a top-pair mass of 345 GeV sits on a larger and differently shaped Standard Model background than previously assumed, and the paper argues its prediction should be used as a conservative input for background modelling and new-physics searches.","feed_headline":"Toponium boosts predicted LHC top-pair background to 12 pb","feed_subtitle":"Near top-pair mass 345 GeV, the Standard Model background rises to 12 pb, versus 8.4 pb in current Monte Carlo models.","key_machinery":"The central object is the NRQCD factorization formula $M_{t\\bar{t}}\\,d\\hat{\\sigma}_{ij\\to T}/dM_{t\\bar{t}} = F_{ij\\to T}(\\hat{s},M_{t\\bar{t}}^2,\\mu_f^2)\\,\\frac{1}{m_t^2}\\mathrm{Im}\\,G^{[1,8]}(M_{t\\bar{t}}+i\\Gamma_t)$, which separates the cross section into a perturbative hard function $F$ and the imaginary part of the non-relativistic Green's function $G$, evaluated at zero distance. The Green's functions solve a Schrödinger equation with the next-to-leading-order QCD potential; the attractive color-singlet solution generates the toponium enhancement, and the repulsive color-octet solution shapes the steeply rising background just above threshold. Threshold logarithms from soft and collinear gluons are resummed in Mellin space to NLL accuracy and matched to NLO fixed order, following the framework of Ref. [8]. The result is a parameter-free Standard Model prediction for $d\\sigma/dM_{t\\bar{t}}$ in the window roughly $|M_{t\\bar{t}}-2m_t|\\lesssim5$ GeV, which the paper proposes to combine additively with fixed-order NNLO predictions above $M_{t\\bar{t}}\\gtrsim350$ GeV.","core_discovery":"The paper's central claim is that QCD bound-state and threshold effects, both part of the Standard Model, produce a clear enhancement in $d\\sigma/dM_{t\\bar{t}}$ for $M_{t\\bar{t}}\\simeq 2m_t$ that is missing from the fixed-order and shower Monte Carlo calculations used in current experimental analyses. Working in non-relativistic QCD, the authors factor the partonic cross section into a hard function times the imaginary part of a non-relativistic Green's function, $F_{ij\\to T}\\times \\frac{1}{m_t^2}\\,\\mathrm{Im}\\,G^{[1,8]}(M_{t\\bar{t}}+i\\Gamma_t)$. The color-singlet Green's function feels an attractive potential and produces the broad 'toponium' resonance near $M_{t\\bar{t}}\\sim 345$ GeV, while the color-octet channels are repulsive and dominate above about 350 GeV. With the hard function resummed to NLL accuracy and matched to NLO fixed order, the integrated cross section in the bin $M_{t\\bar{t}}\\in[340,350]$ GeV is 12 pb for the central scale choice, versus 8.4 pb from the Monte Carlo prediction used in the experimental analysis. The paper further argues that because the missing NNLO and subleading-power corrections are expected to be positive and of order 10%, the NLO+NLL result acts as a lower bound on the exact Standard Model prediction in the threshold region.","pith_inferences":["If the lower-bound property survives a direct test in top-pair production, the statistical significance of the excess near $M_{t\\bar{t}}\\sim345$ GeV would likely drop, because part of the apparent signal would be absorbed into the larger Standard Model background; re-running the experimental fit with this background shape would quantify the shift.","The same machinery could be applied to the spin-correlation and angular observables used to characterise the excess; the predicted background angular distribution would differ from the current Monte Carlo one, giving an additional discriminating test.","A direct verification of the bounding property needs no new data: computing the exact next-to-leading-power threshold terms at NLO in the $gg$ and $q\\bar q$ channels and comparing them with the NLL-resummed result would settle whether the 'lower bound' wording is justified in every channel.","With more LHC data, the 340–350 GeV bin will be populated densely enough that the 12 pb versus 8.4 pb difference may become visible as a pure rate measurement, provided systematic uncertainties on the background shape can be controlled below about 20%."],"forward_implications":["The Standard Model background in the $M_{t\\bar{t}}\\in[340,350]$ GeV window is about 12 pb at NLO+NLL, roughly 40% above the 8.4 pb Monte Carlo estimate, so experimental excesses in this bin must be reinterpreted against the larger background.","The prediction provides a lower bound on $d\\sigma/dM_{t\\bar{t}}$ near threshold, because missing NNLO corrections to the Green's functions and hard functions, and the omitted next-to-leading-power terms, are both estimated to increase the cross section by about 10%.","Above $M_{t\\bar{t}}\\simeq350$ GeV the NRQCD approach ceases to be valid, so fixed-order NNLO and the Monte Carlo description remain the appropriate tools there; the paper recommends a transition region beginning around that mass.","Models that add a toponium-like pseudoscalar resonance to Monte Carlo samples, as used by the experimental analysis, give different peak shape, width, and height; the full NRQCD prediction is a more complete Standard Model template for the bound-state contribution."],"supporting_citations":[{"why":"It defines the experimental excess near twice the top-quark mass and the Run 2 settings (PDF set, scales, Monte Carlo background) that this calculation matches.","marker":"[1]"},{"why":"It supplies the NRQCD factorization setup, the Green's-function machinery, and the matching prescription that this letter updates.","marker":"[8]"},{"why":"It presents the alternative effective-field-theory treatment whose concerns about threshold resummation are addressed and whose region of validity is disputed.","marker":"[10]"},{"why":"It provides the simplified resonant-state model used for toponium in experimental analyses, which this paper argues is incomplete and compares against.","marker":"[11]"},{"why":"It provides the NLO hard functions for quarkonium production in the partonic channels used here.","marker":"[17]"},{"why":"It provides the NNLO fixed-order invariant-mass predictions used for the high invariant-mass matching region.","marker":"[18]"},{"why":"It supplies the parton distribution functions used in the numerical predictions.","marker":"[22]"},{"why":"It establishes the leading-power/next-to-leading-power lower-upper bound property at N4LL for deep-inelastic scattering and Higgs production, which the paper transfers to top-pair production.","marker":"[23, 24]"},{"why":"It provides higher-order electron-positron top-pair threshold results used to estimate that NNLO corrections increase the color-singlet Green's function by about 10%.","marker":"[26]"},{"why":"It supplies the Monte Carlo generator predictions that serve as the comparison baseline for the experimental background.","marker":"[27–29]"}],"fun_headline_variants":["QCD bound states lift top-pair rate to 12 pb in threshold bin","Toponium resonance raises predicted LHC top-pair background","NLO+NLL lower bound boosts top-pair mass bin to 12 pb","Threshold resummation lifts top-pair cross section near 345 GeV"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the leading-power/next-to-leading-power bounding property, proven to N4LL accuracy for deep-inelastic scattering and Higgs production in Refs. [23, 24], carries over to top-quark-pair hadroproduction in the three channels considered; if that transfer fails, the paper's statement that the NLO+NLL result is a lower bound on $d\\sigma/dM_{t\\bar{t}}$ is unsupported even though the NLL calculation itself might still be correct.","fun_headline_variants_meta":{"raw":{"variants":["QCD bound states lift top-pair rate to 12 pb in threshold bin","Toponium resonance raises predicted LHC top-pair background","NLO+NLL lower bound boosts top-pair mass bin to 12 pb","Threshold resummation lifts top-pair cross section near 345 GeV"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001286,"raw_usage":{"total_tokens":5246,"prompt_tokens":927,"completion_tokens":4319,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":543,"completion_tokens_details":{"reasoning_tokens":4239}},"tokens_in":543,"tokens_out":4319,"duration_ms":26458,"temperature":1.0,"reasoning_tokens":4239,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:21:43.054083+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The transfer of the LP/NLP bounding property can be tested directly: compute the exact NLO (or NNLO) threshold corrections including next-to-leading-power terms for the three channels $gg\\to {}^1S_0^{[1]}$, $gg\\to {}^1S_0^{[8]}$, and $q\\bar q\\to {}^3S_1^{[8]}$ in Mellin space and check whether the NLO+NLL resummed result lies below them in every channel. A single channel where it lies above would falsify the lower-bound claim; alternatively, a measurement of the 340–350 GeV bin at the LHC with total uncertainty below about 20% would discriminate 12 pb from 8.4 pb directly.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It defines the experimental excess near twice the top-quark mass and the Run 2 settings (PDF set, scales, Monte Carlo background) that this calculation matches."}],"review_version":1}