{"id":"eabe6b0e-b527-4db2-b7db-79eea321af90","arxiv_id":"2412.19694","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Monte Carlo generators predict that nonperturbative corrections for Z+jet production depend strongly on the scattering angle y* and are process-dependent, unlike dijet production, suggesting the corrections mix perturbative and nonperturbative physics.","lead":"This simulation study finds that the nonperturbative corrections used to compare LHC data with perturbative QCD predictions behave differently for Z+jet events than for dijet events: the Z+jet corrections grow with the scattering angle y* while dijet corrections stay flat. The authors argue part of what is labeled nonperturbative actually comes from perturbative effects and propose a new triple-differential measurement to test this.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The y* trend in Z+jet NP corrections is established only through the smoothing fit of Eq. (7); raw binned points are never shown, so a fit artifact could drive the central claim.","rationale":"The paper is a Monte Carlo study whose central claim—that NP correction factors for Z+jet production depend strongly on y* but barely on yb—is shown only through smoothed curves. The reader's weakest_assumption already includes the smoothing function as a structural premise, and my concern sharpens this: without the raw binned ratios, there is no way to know whether the y* ordering is real or induced by the fit. The appendix itself flags the low-edge slope risk, so this is not an external objection but a direct reading of the manuscript's own caution. I do not question the authors' honesty or the value of the study: the two-generator agreement and the persistence of y* dependence across perturbative orders (including multijet merging) provide independent support that the qualitative effect is unlikely to be a pure statistical fluke. However, the quantitative claim that standard correction factors 'strongly depend on y*' is exactly the kind of statement that should be verified against the unsmoothed points before being used to motivate a new measurement. The proposed test—displaying raw points and refitting with a different functional form—would settle whether the headline observation is robust. Since the reader's conditional verdict already demands the unsmoothed MC points and addresses the omitted bin, my read does not change the verdict; it only reinforces the condition.","tokens_in":22590,"tokens_out":6353,"duration_ms":68987,"concrete_test":"Regenerate the Z+jet HERWIG7 and SHERPA samples using the released MCRun configurations and compute the raw C_NP per (yb,y*) bin with statistical uncertainties, without applying Eq. (7). For each pT_Z bin in the overlap range, test the y* ordering across the five y* bins at central yb (e.g., a chi-square or Spearman rank correlation). Then refit the raw points with at least one alternative functional form, such as c + a/x^b or a cubic spline, and compare the extracted y* ordering and magnitude. If the raw points show no significant y* trend, or if the alternative fit reverses or weakens it, the central claim should be downgraded to a fit-dependent observation rather than a robust MC prediction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central observation—that C_NP in Z+jet production increases strongly with y* at fixed pT_Z while showing no yb dependence—is presented exclusively as fits of Eq. (7), f(x) = a ln(x/GeV)^b + c, to binned MC ratios. The unsmoothed ratios are never displayed, and Appendix A explicitly warns that when the low-scale rise is represented by a single bin at the low edge, statistical fluctuations can produce exaggerated slopes. Since the y* separation is largest at the lowest pT_Z bin, a few downward-fluctuating points in low-y* bins or upward-fluctuating points in high-y* bins could induce the entire observed ordering after the log-power fit. Moreover, the omitted outermost y* bin (2.0 ≤ y* < 2.5) in the UE analysis of Section 5.2 is exactly the bin where the effect is claimed strongest; without it, the trend in the main text could be biased. Because the paper's conclusion that conventional correction factors are process-dependent and partly perturbative rests on this y* dependence, the claim is load-bearing on the unsmoothed data and the stability of the fit choice.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper investigates nonperturbative (NP) corrections, defined in Eq. (6) as the ratio of MC cross sections with hadronisation and MPI switched on versus off, for triple-differential dijet and Z+jet production at the LHC. Using HERWIG7 and SHERPA at LO, NLO, and merged ME accuracy, the authors fit Eq. (7), f(x)=a·ln(x/GeV)^b+c, to the binned correction factors and report that for Z+jet production C_NP depends strongly on y* but barely on yb, while dijet production shows no such dependence. They trace the effect mainly to MPI and, through a UE analysis, argue that part of the activity is of perturbative origin. The paper concludes that conventional NP correction factors may not be entirely NP in origin and are not universal across hard processes, and it proposes a triple-differential measurement of the underlying event in Z+jet production.","tokens_in":22832,"tokens_out":6206,"duration_ms":59514,"significance":"If the results hold, they are relevant for precision QCD because standard NP correction factors used in alpha_s and PDF fits would be process-dependent and could contain perturbative contamination. The paper's strengths are the systematic comparison of two independent generators, several perturbative orders (LO, NLO, merged), and an explicit on/off decomposition of hadronisation and MPI, plus a reproducible workflow via the MCRun framework. The authors are appropriately cautious in stating that whether this behaviour is realized in nature remains an open question. However, the central qualitative claim is currently established only within the default MC models, and one key piece of evidence—the unsmoothed C_NP values—is not shown.","major_comments":[{"comment":"The central observation—that C_NP for Z+jet increases strongly with y* at fixed pT_Z—is presented exclusively through the fitted curves of Eq. (7). The unsmoothed binned ratios are never shown, so the reader cannot assess whether the separation of the curves at low pT_Z is driven by the fit function rather than by the underlying MC points. Appendix A itself warns that when the low-scale rise is represented by only one bin at the low edge, statistical fluctuations may lead to exaggerated slopes. Because the y* separation is largest at the lowest pT_Z bins, the authors should display the raw C_NP values with statistical uncertainties for the nine central (y*, yb) bins and demonstrate stability of the fitted trends under removal of the lowest pT_Z bin, variation of the fit function, and changes of the fit range.","section":"Sec. 3, Eq. (7), Appendix A"},{"comment":"The UE analysis omits the outermost y* bin (2.0 ≤ y* < 2.5), citing large statistical fluctuations, but this is exactly the bin where the y* dependence of C_NP in Fig. 4 appears strongest. The consistency argument that the UE activity exhibits the same y* dependence as C_NP is therefore incomplete at the most extreme y*. The authors should either include this bin with properly inflated uncertainties or merge it with the adjacent bin and show that the trend persists.","section":"Sec. 5.2"},{"comment":"The dijet and Z+jet analyses differ in kinematics: dijet uses <pT>1,2 with bins starting at 47 GeV, while Z+jet uses pT_Z starting at 25 GeV, and the jet pT thresholds differ (30/25 GeV for dijet versus 20 GeV for Z+jet). The strongest y* dependence in Z+jet is found at low pT_Z, a region not covered by the dijet comparison. This raises the question whether the observed process non-universality is a genuine property or a kinematic effect. A matched comparison in the overlapping pT range, or a reanalysis of dijet with lower thresholds, would strengthen the non-universality claim.","section":"Sec. 2 and Sec. 3"},{"comment":"All results are obtained with the default tunes of HERWIG7 and SHERPA; no variation of MPI or hadronisation parameters is shown. Since the paper attributes a substantial part of the effect to MPI modelling, the robustness of the y* dependence and of the LO/NLO decrease across the available tune variations should be demonstrated. The caveat that realisation in nature remains an open question is appropriate, but without a tune scan the model dependence of the quoted trends is not quantified.","section":"Sec. 3 and Sec. 4"}],"minor_comments":[{"comment":"The notation a·ln(x/GeV)^b is ambiguous; it should be written as a·(ln(x/GeV))^b to avoid confusion over whether the power b acts on the logarithm or on x/GeV.","section":"Sec. 3, Eq. (7)"},{"comment":"The fitted values of a, b, and c are not tabulated, although the text states that c is compatible with unity; providing the fit parameters (or a summary table) would allow the reader to check this compatibility quantitatively.","section":"Sec. 3"},{"comment":"The sign convention in x_{1|2} = m_{1,2}/(sqrt(s)·exp(±yb)) should be clarified by stating explicitly which momentum fraction corresponds to the plus sign and which to the minus sign.","section":"Sec. 2, Eq. (3)"},{"comment":"The sentence 'This would lead to uncorrelated events from MPI and to a strong growth of charged particle multiplicity with the number of additional hard events' is unclear and should be rephrased to separate the expected behaviour without colour reconnection from the role played by colour reconnection.","section":"Sec. 4"},{"comment":"The remark that the mass binning 'has not been extended to include results at lower pT' is ambiguous; please clarify whether the mass binning simply starts at a higher threshold than the <pT> binning and why this choice was made.","section":"Appendix A, Table 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is clearly written and the authors are appropriately cautious in their conclusions. The main methodological concern is that the central trend is only visible after smoothing with Eq. (7) and the raw binned points are never displayed; this is a fixable presentation and analysis gap, not a fundamental flaw. I would not reject, but the revision should include the unsmoothed C_NP values and a stability analysis of the fits, plus a treatment of the omitted high-y* bin in the UE section. The comparison between dijet and Z+jet in matched kinematics would also materially strengthen the non-universality claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Main thing to know: this is a solid MC study, not a discovery. The new observation is that conventional NP correction factors, defined as the ratio particle-level/parton-level, depend strongly on y* for Z+jet but not for dijet, and that part of that dependence survives even when MPI is switched off. If it holds in data, it matters for alpha_s and PDF fits. The authors are appropriately careful: the abstract says the non-universal behaviour may not be realized in nature.\n\nCredit where due. The study is well set up: two generators (HERWIG7, SHERPA), LO/NLO and merged MEs, jet radii 0.4 and 0.8, explicit separation of hadronisation vs MPI, all 15 (yb,y*) bins in the appendix, and MCRUN/Rivet configurations for reproduction. The decomposition is particularly useful: the y* trend is traced to MPI, with hadronisation flat, and the UE analysis shows y*-dependent activity even with MPI off, pointing to perturbative contamination. The claim is not circular; the y* dependence is a generator output, not an input.\n\nSoft spots. The stress-test concern is real. In the main plots, only fitted curves from Eq. (7) are shown; raw binned ratios never appear. The function is ad hoc, and the appendix itself warns that a steep low-scale rise represented by a single bin at the low edge can give exaggerated slopes. Since the biggest y* separation sits at the lowest pTZ and in the outermost y* bin, a fit artifact could distort the ordering. The UE section omits 2.0 <= y* < 2.5 because of statistics, which is exactly the bin where the effect looks strongest. There is no systematic variation of tunes or colour-reconnection models, so the quantitative size of the effect is not pinned down. None of this kills the claim: two generators agree, the y* ordering is consistent across ME orders, and the UE plots show the trend in a different observable before any fit. But it does mean the headline should be carried by unsmoothed points and stability checks, not by the fitted curves. The citations look appropriate, and the related inclusive-jet work is properly cited.\n\nBottom line: useful paper for precision-QCD and MC people. Deserves a serious referee. I would send it out, and ask for the raw points, an alternative smoothing or no-smoothing view, a tune variation, and the outermost bin restored if possible.","headline":"A plausible and genuinely new MC result on process-dependent nonperturbative corrections, but the central y* trend is currently shown only through an ad hoc smoothing fit; the paper deserves review with the raw points on the table.","tokens_in":23387,"tokens_out":2495,"would_cite":true,"duration_ms":29522,"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":"Nonperturbative corrections for Z+jet events grow with the centre-of-mass scattering angle $y^*$ but barely with the boost $y_b$, and are not universal across hard processes.","keywords":["nonperturbative corrections","dijet production","Z+jet production","triple-differential cross section","underlying event","hadronisation and MPI","Monte Carlo event generators","strong coupling and PDFs"],"falsifier":"Measure the underlying-event activity in Z+jet events at the LHC in bins of $p_T^Z$, $y_b$, and $y^*$: the paper predicts the charged-particle transverse-momentum sum in the transverse azimuthal region rises with $y^*$ and stays flat in $y_b$, so data showing the opposite dependence would falsify the central observation.","tokens_in":22388,"feed_emoji":"⚛️","tokens_out":9901,"duration_ms":83994,"temperature":0.7,"pith_summary":"Precision determinations of the strong coupling constant and the proton's parton distribution functions rely on dijet and Z+jet cross sections that are computed in perturbative QCD but must be corrected for nonperturbative effects such as hadronisation and the underlying event. Using the HERWIG7 and SHERPA event generators, this paper computes those corrections in a triple-differential binning in the scale variable, the longitudinal boost $y_b$, and the centre-of-mass scattering angle $y^*$. For Z+jet events the correction factors rise toward small $p_T^Z$ and this rise becomes stronger as $y^*$ grows, while they stay nearly flat in $y_b$; for dijet production neither angular dependence appears. The paper concludes that these conventionally defined corrections are not entirely nonperturbative in origin and are not universal between hard processes, and it proposes a triple-differential underlying-event measurement at the LHC to test whether the angular behaviour is realised in data.","feed_headline":"Z+jet nonperturbative corrections grow with scattering angle","feed_subtitle":"Dijet factors stay flat, so standard QCD correction factors need process-specific re-examination.","key_machinery":"The load-bearing object is the ratio $C_{\\mathrm{NP}}=\\sigma_{\\mathrm{ME+PS+Had+MPI}}/\\sigma_{\\mathrm{ME+PS}}$ taken in each $(y_b,y^*)$ bin and expressed as a function of $\\langle p_T\\rangle$ or $p_T^Z$; it defines what the paper means by a nonperturbative correction. The analysis separates this ratio into a hadronisation-only factor $C_{\\mathrm{Had}}$ and an MPI-only factor $C_{\\mathrm{MPI}}$, and then examines underlying-event observables — the charged-particle transverse-momentum sum in the towards, away, and transverse azimuthal regions — in the same binning. The smoothing function $f(x)=a\\ln(x/\\mathrm{GeV})^b+c$ is fitted to the correction factors to suppress statistical fluctuations; the fitted constant $c$ is close to unity, consistent with corrections fading at high energy.","core_discovery":"The central claim is that the nonperturbative correction factor $C_{\\mathrm{NP}}$, defined for each bin as the ratio of the particle-level cross section to the same generator's prediction with hadronisation and multiple-parton interactions switched off (Eq. 6), behaves differently for the two processes. In Z+jet production, $C_{\\mathrm{NP}}$ deviates from unity most at low $p_T^Z$, and the deviation grows systematically with the scattering-angle variable $y^*$ but hardly changes with the boost $y_b$; upgrading the matrix element from LO to NLO, or to multijet-merged NLO, reduces the size of the corrections. In dijet production, by contrast, the corrections are close to unity at high $\\langle p_T\\rangle$ and show no significant $y^*$ or $y_b$ dependence. Since the same $y^*$ trend appears in underlying-event observables even with the MPI model turned off, and since the effect shrinks with more complete perturbative input, the paper concludes that the conventional corrections may contain perturbative contamination and are not universal across hard processes.","pith_inferences":["If LHC data confirm the $y^*$ trend, the effect could bias global fits that apply universal NP corrections, shifting $\\alpha_s$ and gluon-PDF determinations at the percent level depending on the kinematic region — a testable, quantitative consequence the paper leaves implicit.","The persistence of the $y^*$ dependence with MPI switched off suggests the transverse azimuthal region is not a clean probe of the underlying event in Z+jet events at high $y^*$; differential UE tuning, not inclusive UE tuning, would be needed to calibrate it.","A natural next generator-level test is to vary the colour-reconnection model or hadronisation parameters and check whether the $y^*$ slope changes; the paper does not perform a full model-variation study.","The same angular-variable decomposition could be applied to W+jet or photon+jet production, where a $y^*$ dependence of NP corrections, if present, would reveal whether the effect is specific to colour-singlet-plus-jet topologies or a generic feature of leptonically tagged processes."],"forward_implications":["If the non-universality is real, precision extractions of $\\alpha_s$ and the gluon PDF from Z+jet data need process-specific nonperturbative corrections rather than factors carried over from dijet or inclusive-jet fits.","The $y^*$ dependence implies that part of what is labelled nonperturbative is actually perturbative radiation from additional jets, so Z+jet event generation for such fits should use multijet merging at NLO, which reduces the corrections.","The paper's proposed triple-differential measurement of the underlying event in Z+jet events would directly test whether the generator-level angular pattern appears in LHC data.","For dijet production, the flatness of the correction factors in $y_b$ and $y^*$ suggests the standard treatment is less affected by this angular issue, though the same statement about perturbative contamination may still apply.","Because the corrections depend on the perturbative order of the matrix element (LO vs NLO vs merged), the numerical value of a nonperturbative correction only makes sense when quoted together with the order and parton-shower setup."],"supporting_citations":[{"why":"supplies the NNLO predictions that define the precision context in which nonperturbative corrections are needed.","marker":"[1–5]"},{"why":"sets the triple-differential $y_b$–$y^*$ binning and phase-space choices used for the dijet analysis.","marker":"[12, 13]"},{"why":"provides the HERWIG7 generator whose LO and NLO samples yield the main correction factors.","marker":"[21, 22]"},{"why":"supplies SHERPA as the independent cross-check generator for the observed trends.","marker":"[23]"},{"why":"earlier evidence that correction factors shift with perturbative order and rapidity, supporting the paper's interpretation.","marker":"[28, 29]"},{"why":"models colour reconnection that ties multiple-parton interactions to hadronisation, justifying the partial decomposition of $C_{\\mathrm{NP}}$.","marker":"[30–32]"},{"why":"defines the underlying-event observables used to probe the angular dependence of the extra activity.","marker":"[36–38]"}],"fun_headline_variants":["Z+jet nonperturbative corrections grow with angle, dijet flat","Nonperturbative QCD corrections differ between Z+jet and dijet","Standard correction factors may contain perturbative contamination","Proposed Z+jet underlying event measurement tests non-universality","Perturbative contamination suspected in nonperturbative corrections"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the hadronisation and multiple-parton-interaction models of HERWIG7 and SHERPA reproduce the real physics of nonperturbative effects closely enough that the predicted $y^*$ dependence in Z+jet events, and its absence in dijets, would appear in LHC data.","fun_headline_variants_meta":{"raw":{"variants":["Z+jet nonperturbative corrections grow with angle, dijet flat","Nonperturbative QCD corrections differ between Z+jet and dijet","Standard correction factors may contain perturbative contamination","Proposed Z+jet underlying event measurement tests non-universality","Perturbative contamination suspected in nonperturbative corrections"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000225,"raw_usage":{"total_tokens":1427,"prompt_tokens":874,"completion_tokens":553,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":490,"completion_tokens_details":{"reasoning_tokens":464}},"tokens_in":490,"tokens_out":553,"duration_ms":6460,"temperature":1.0,"reasoning_tokens":464,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:57:38.570003+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the underlying-event activity in Z+jet events at the LHC in bins of $p_T^Z$, $y_b$, and $y^*$: the paper predicts the charged-particle transverse-momentum sum in the transverse azimuthal region rises with $y^*$ and stays flat in $y_b$, so data showing the opposite dependence would falsify the central observation.","supporting_citations":[],"review_version":1}