{"id":"8b7e0706-8ee4-4aa3-bc03-979565f2b03b","arxiv_id":"1301.7310","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors compute the NNLO QCD corrections to inclusive jet and dijet production in the pure-gluon channel at leading colour, the first NNLO jet observable at a hadron collider.","lead":"Physicists report the first next-to-next-to-leading-order QCD calculation of a jet observable at a hadron collider, carried out for the pure-gluon channel at leading colour. It is a methodological milestone on the road to precision jet predictions for LHC measurements.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The finite NNLO distributions are shown to be finite but not independently validated: pole cancellation is cited from Refs. [40,41], and no NLO-limit or subtraction-parameter check confirms the finite remainder is correct.","rationale":"This is a well-structured letter reporting a genuine first: an NNLO calculation of a jet observable at a hadron collider, in the gluonic channel at leading colour, implemented in NNLOJET. The claim in abstract:s1 is plausible and the technical scaffolding (Eq. 1, the wedge decomposition, the four figures) is coherent. The paper is appropriately careful in the conclusion: it states that the calculation is 'gluons-only' in the sense of matrix elements (footnote [47]) and that the all-gluon process 'can therefore not be directly compared with experimental data.' Those acknowledgements are credited and are not the concern. The load-bearing concern is the absence of independent validation of the finite NNLO remainder. The pole cancellation is demonstrated analytically in Refs. [40,41], and Ref. [41] is an arXiv preprint (1211.2710) without a journal reference at the time of writing — a legitimate flag under the citation rules, though the citation does appear to contain the result it is used for. The deeper issue is that pole cancellation tests only the divergent part; the finite part, which is what Figs. 1-5 show, is not independently checked. A finite but wrong result would be indistinguishable from a correct one on the evidence presented. The reader's weakest_assumption identifies the right general area (the subtraction's correctness is not independently validated) but frames it primarily as a citation issue rather than a finite-part validation issue; hence 'partial' agreement. The reader's other concerns are real but minor: the leading-colour truncation appears only in the conclusion while the abstract says 'purely gluonic channel' (a presentational overstatement, though the paper does acknowledge it); no Monte Carlo uncertainties are quoted on the 15-25% k-factors (a presentational gap, not a correctness issue); and no code or data are released (common for 2013). None of these breaks the central claim, which is a methodological milestone. I recommend CONDITIONAL, unchanged from the reader. The concrete test — varying the unphysical subtraction parameters and checking the NLO limit — would settle whether the finite remainder is validated. If the test passes, the paper's claim is well supported; if it fails, the claim weakens substantially.","tokens_in":9466,"tokens_out":11432,"duration_ms":214903,"concrete_test":"Run NNLOJET in the all-gluon channel and vary the unphysical subtraction parameters: the boundaries of the 6 three-parton and 30 four-parton wedges and the angular-rotation parameters used to generate multiple phase-space configurations. If the NNLO cross section is invariant to within Monte Carlo error, the subtraction's finite remainder is validated; if it shifts, the subtraction terms do not fully cancel and the central claim weakens. As a complementary check, confirm that the O(αs^3) expansion of the NNLO result reproduces the published all-gluon NLO jet cross sections of Refs. [42-44] for the same cuts (anti-kT R=0.7, |y|<4.4, pT>80 GeV, √s=8 TeV).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim — 'first NNLO calculation of a massless jet observable at hadron colliders' (abstract:s1) — rests on the finite NNLO distributions in Figs. 1-5 being correct, not merely finite. Eq. (1)'s three brackets are finite by construction if the antenna subtraction is correct, and the analytic pole cancellation is demonstrated in Refs. [40,41]. But finiteness is necessary, not sufficient: an error in the finite part of the integrated antenna functions, or in the assembly of the two-loop, one-loop, and double-real pieces, would produce a finite but wrong result that the pole check cannot detect. The paper provides no numerical validation of the finite remainder: no demonstration that the NLO limit reproduces the published all-gluon NLO jet results of Refs. [42-44], no check that the NNLO cross section is invariant under variation of the unphysical subtraction parameters (the 6 three-parton and 30 four-parton wedge boundaries and the angular-rotation parameters), and no independent NNLO cross-check (none was available in 2013). The smooth curves in Figs. 1-5 are consistent with correctness but do not establish it. This is the load-bearing gap because the entire claim depends on the finite numbers being right, and the paper's evidence for that is the pole cancellation plus the plausibility of the curves. The paper does acknowledge that the all-gluon process 'can therefore not be directly compared with experimental data' and that the calculation is 'gluons-only' in the sense of matrix elements (footnote [47]), which is the correct framing; the unresolved issue is the internal validation of the finite remainder.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The manuscript reports the first next-to-next-to-leading order (NNLO) QCD calculation of a jet-defined observable at a hadron collider, restricted to the purely gluonic channel at leading colour. Using antenna subtraction, the authors combine two-loop virtual, one-loop real, and double-real gluonic matrix elements inside the parton-level generator NNLOJET, with the three subtraction terms of Eq. (1) rendering each of the four-, three-, and two-parton integrations finite. They present, for pp collisions at 8 TeV with the anti-kT algorithm (R = 0.7), the inclusive jet pT spectrum (Fig. 1), its scale dependence in one pT bin (Fig. 2), a double-differential pT and rapidity spectrum (Fig. 3) with the corresponding k-factors (Fig. 4), and the exclusive dijet invariant-mass distribution (Fig. 5). The headline quantitative results are NNLO/NLO k-factors of roughly 15-25% and a substantial reduction of the factorization and renormalization scale dependence relative to NLO. The authors state explicitly that the all-gluon channel cannot be compared directly with data and that the calculation is 'gluons-only' in the sense that only gluonic matrix elements enter.","tokens_in":9736,"tokens_out":4387,"duration_ms":79733,"significance":"If the finite remainder is correct, this is a genuine milestone: the first demonstration that a jet-defined hadron-collider observable can be computed at NNLO with all infrared poles cancelled locally, and the first quantitative estimate of the size of NNLO corrections to jet production. The calculational machinery (antenna subtraction with initial-state radiators, wedges, and rotation-averaged phase space) is a substantial and reusable asset, and the paper is candid that the all-gluon channel is a methodological stepping stone rather than a phenomenological prediction. The result is a first-principles perturbative calculation with no tuned parameters, and the authors are appropriately explicit about the scope of what can and cannot be concluded from it. The credit due here is for opening a programme, and the manuscript states that clearly.","major_comments":[{"comment":"No Monte Carlo statistical uncertainty is quoted anywhere in the letter. The distributions in Figs. 1, 3, and 5 carry no error bars, the scale bands in Fig. 2 carry no error estimate, and the quantitative claims in the text (a 15-25% NNLO/NLO k-factor in Fig. 1, 25% at low pT falling to 12% at high pT in Fig. 4, and 15-20% in Fig. 5) are stated without an assessment of whether they are resolved above integration noise. This matters most in the highest-pT bins, where the cross section falls by several orders of magnitude and where a multi-channel generator with 6 three-parton and 30 four-parton wedges is most vulnerable to poor sampling. Since the central quantitative message of the letter is the size of the NNLO correction, the reader needs to know the numerical accuracy of each bin. Please add statistical error bars (or a tabulated per-bin uncertainty) to Figs. 1-5 and state the achieved integration accuracy and the number of events generated.","section":"Figs. 1-5 and the quoted k-factors","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The finiteness of each bracket in Eq. (1) is asserted by citation to Refs. [40,41], and the letter offers no numerical validation that the finite remainder is correct rather than merely finite. An error in the finite part of an integrated antenna function, or in the assembly of the two-loop, real-virtual, and double-real pieces, would leave the pole cancellation intact and would not be visible in the smooth curves of Figs. 1-5. Two cheap and standard cross-checks are absent: (i) a demonstration that the code reproduces the published all-gluon NLO jet results of Refs. [42-44], which the manuscript itself cites as the historical precedent for establishing methodology in this channel; and (ii) a demonstration that the NNLO cross section is invariant under variation of the unphysical subtraction parameters, namely the wedge boundaries and the angular-rotation parameters used to generate the multiple phase-space configurations. If such checks were performed, they should be reported; if they were not, at minimum the manuscript should state which cross-checks were done and what residual validation is deferred to the companion work.","section":"Eq. (1) and the surrounding discussion of the subtraction","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The abstract claims 'the first NNLO calculation of a massless jet observable at hadron colliders' and describes the result as 'in the purely gluonic channel', but the leading-colour restriction appears only in the Conclusions ('the purely gluonic channel at leading colour'). The abstract and Introduction should carry the leading-colour qualifier explicitly, and the 'first NNLO calculation' phrasing should be sharpened to make clear that this is the first NNLO calculation of a jet observable in the all-gluon channel at leading colour, not a complete NNLO prediction for any jet observable. As written, a reader of the abstract alone could reasonably infer a complete result. The body of the letter is careful about this; the abstract should match it.","section":"Abstract (abstract:s1) and Introduction versus the Conclusions","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"Fig. 5 uses mu = 2 pT1 while Figs. 1-4 use mu = pT1, and no motivation is given for the different default. Please state why the dijet-mass observable uses a different central scale, or use a uniform choice throughout.","section":"Fig. 5 and the scale choice","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The scale-dependence plot is shown for a single pT bin (80-97 GeV), while the text states that 'the scale dependence of other pT and y slices is also reduced' without showing them. Either show a representative second slice or soften the generalization to a statement about the bin displayed.","section":"Fig. 2 and the accompanying text","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Ref. [41] (arXiv:1211.2710) is listed without a journal reference, and it is the source of the analytic pole cancellation on which Eq. (1) rests. Please update the citation if the companion paper has since been published.","section":"Reference [41]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The labels S, T, and U on the subtraction terms d sigma^{S,T,U} are not defined in the text; the reader must infer from the equation that S accompanies the four-parton, T the three-parton, and U the two-parton channel. A one-sentence definition would help.","section":"Eq. (1)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sentence 'We expect similar conclusions when including the processes involving quarks' is a speculation about channels not computed here. It should be flagged as such, since the quark channels could plausibly change both the k-factor and the scale-reduction picture.","section":"Conclusions","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text describes jets with |y| < 4.4 as accepted 'at central rapidity'; 4.4 is the forward acceptance boundary, not a central region. Please reword.","section":"Jet selection wording","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Fig. 1 is captioned as a 'transverse energy distribution' but the axis and text use pT, and Fig. 3 uses 'ET > 80 GeV' where the text uses pT > 80 GeV. Please make the notation uniform across figures and text.","section":"Fig. 1 and Fig. 3 captions","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The PACS line reads '13.87.Ce,12.38Bx'; the second entry appears to be missing a period and should presumably read 12.38.Bx.","section":"PACS numbers","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"This is a landmark letter and the physics is sound; my recommendation is minor revision because the requested changes are quantitative reporting and precision of claim, not corrections to the calculation. The one point I would press the authors on is the absence of any statistical uncertainty on the k-factors and scale bands, since those numbers are the paper's headline. The heavy reliance on the authors' own Refs. [39-41] is normal for a cumulative programme and not a concern in itself, but the editor may wish to confirm that the companion paper [41] is published or in press before this letter appears, given that the pole cancellation is cited rather than re-derived here."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Short version: this is the real thing. First NNLO calculation of a jet observable at a hadron collider, and it holds up as a methodological milestone. The caveats are about how much of the result you can take quantitatively, not about whether the calculation was done at all. What's new: the subtraction machinery is theirs (Refs [39-41]), but the generator implementation, the wedge decomposition (6 three-parton, 30 four-parton wedges with angular rotations to force local convergence), and the first finite NNLO distributions for inclusive jet pT, its scale dependence, the double-differential pT/|y| spectrum, and dijet mass at 8 TeV are not in the prior literature. The scale-band collapse in Fig. 2 is the headline physics result and it is convincing as far as it goes. Soft spots, in order of weight. First, Ref [41], which carries the analytic pole cancellation, is cited as arXiv:1211.2710 with no journal reference — unpublished when this was written. Everything finite in Eq. (1) depends on it. Not fatal (it did appear in JHEP), but a letter leaning this hard on an unpublished companion should say so. Second, the stress-test point lands: finiteness is not correctness. There is no NLO-limit check against Refs [42-44], no demonstration that the cross sections are invariant under the wedge boundaries and rotation parameters, and no independent NNLO cross-check (none existed in 2013). An error in the finite part of the integrated antenna functions would give smooth, finite, wrong curves, and the pole check cannot see it. I would want at least the NLO limit shown before quoting these numbers. Third, no Monte Carlo uncertainties on any figure; the 15-25% k-factors are large enough that this probably does not change the story, but it should be stated. Fourth, leading colour appears only in the conclusion while the abstract says 'purely gluonic channel' without qualification — minor, but worth tightening. Credit where due: the authors are explicit that the all-gluon channel cannot be compared to data, and footnote [47] is honest about 'gluons-only' meaning matrix elements. Who should read it: anyone doing jet phenomenology, alpha_s extraction, or NNLO subtraction. It is a letter, so the details live elsewhere, but the result sets the agenda for the next couple of years. Recommendation: send it out. The claim is right and the method is sound; referees should push on validation of the finite remainder and on the leading-colour qualifier.","headline":"A genuine first — NNLO jet production at a hadron collider — but the finite numbers rest on an unpublished companion and no validation of the finite remainder.","tokens_in":10457,"tokens_out":2055,"would_cite":true,"duration_ms":37176,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["13.87.Ce","12.38.Bx"],"model":"selfhosted/deepseek-v4.1-flash","headline":"The paper reports the first next-to-next-to-leading-order QCD calculation of a jet observable at a hadron collider, carried out in the purely gluonic channel with antenna subtraction, and shows the residual scale uncertainty shrinking by a ","keywords":["NNLO QCD","jet production","antenna subtraction","hadron colliders","infrared singularities","parton-level event generator","dijet invariant mass","strong coupling"],"falsifier":"Run the generator with the NNLO terms switched off and compare the resulting all-gluon jet cross sections against the published NLO all-gluon results of Refs. [42-44]; a mismatch, or a residual uncancelled pole appearing in any of the three channels of Eq. (1), would undercut the claim that the subtraction is complete for jet observables.","tokens_in":10774,"feed_emoji":"⚛","tokens_out":3505,"duration_ms":76142,"temperature":0.7,"pith_summary":"The paper sets out to compute the next-to-next-to-leading order (NNLO) correction in QCD to jet production at a proton-proton collider, restricted for now to the purely gluonic channel at leading colour. Jets are the workhorse observable of the LHC: they constrain the gluon content of the proton and give one of the cleanest handles on the strong coupling $a_s$, yet the theory error from missing higher orders has been larger than the experimental error, which is what has kept jet data out of NNLO fits. The authors implement the calculation in a parton-level event generator, NNLOJET, whose three integration channels each pair a double-real, real-virtual or two-loop virtual matrix element with a matching antenna subtraction term, so that each bracket in Eq. (1) is finite. They then present the inclusive jet $p_T$ spectrum, its scale dependence, a double-differential $p_T$ and rapidity spectrum, and the dijet invariant mass at $/sqrt{s}=8$ TeV. If the cancellation holds, this is the first demonstration that a jet-defined observable at a hadron collider can be computed at NNLO.","feed_headline":"First second-order QCD prediction for hadron-collider jets","feed_subtitle":"The all-gluon calculation cuts scale dependence and lifts jet rates 15-25 percent above next-to-leading order.","key_machinery":"Antenna subtraction. An antenna function is a compact analytic object that captures all the unresolved radiation emitted between a pair of hard radiator partons, so that subtracting it from the corresponding matrix element leaves a finite remainder that can be integrated numerically, while the subtraction term itself can be integrated analytically and added back to the virtual contributions to cancel the explicit infrared poles. At NNLO the scheme needs antenna functions with up to two unresolved partons at tree level and one unresolved parton at one loop, and for hadron collider observables one or both radiators may sit in the initial state. The practical carrier is the phase-space decomposition: the three-parton channel is split into 6 wedges and the four-parton channel into 30, each wedge containing only a subset of the infrared singular regions, with several configurations generated inside each wedge by rotating unresolved pairs about their common axis so that the subtraction term converges locally to the matrix element. Symmetry of the all-gluon final state makes many wedges identical and speeds up the evaluation.","core_discovery":"The claim is that a massless jet observable at a hadron collider has now been computed at NNLO in QCD, in the all-gluon channel. Combining the two-loop virtual $2/rightarrow 2$ corrections, the one-loop virtual $2/rightarrow 3$ corrections and the tree-level $2/rightarrow 4$ corrections with their respective antenna subtraction terms, the authors obtain finite distributions for proton-proton collisions at $/sqrt{s}=8$ TeV with the anti-$k_T$ algorithm at $R=0.7$, $|y|<4.4$ and $p_{T1}>80$ GeV, using the MSTW08NNLO gluon distribution. For the central scale $/mu=p_{T1}$ they find the NNLO/NLO $k$-factor to be roughly flat in $p_T$, a 15-25% increase over NLO, and 15-20% for the dijet invariant mass at $/mu=2p_{T1}$; the double-differential $k$-factors run from about 25% at low $p_T$ to 12% at high $p_T$ across the three central rapidity slices shown. The scale dependence of the inclusive cross section is reported as vastly reduced. The authors state plainly that the all-gluon channel alone cannot be compared with experimental data, and that its purpose is to establish the method and gauge the potential size of NNLO effects.","pith_inferences":["The quoted 15-25% $k$-factor is not the full NNLO correction to dijet production: it is the gluonic channel only, and quark-initiated subprocesses dominate at high $p_T$ and at large rapidity, so the true shift may differ in both normalisation and shape from what is shown here.","If the scale band genuinely shrinks as far as the figures indicate, the binding constraint on jet-based determinations of $a_s$ would move from scale variation to parton distributions and non-perturbative corrections, which changes what the next round of measurements should target.","An independent implementation of the same all-gluon channel by a different subtraction method would be a cheap and decisive cross-check, and would also test whether the wedge-based local convergence is doing any hidden work.","A natural next test is whether the generator, run in its NLO limit, reproduces the published all-gluon NLO jet results; agreement there would isolate any residual issue to the genuinely new two-loop and double-real pieces."],"forward_implications":["The authors state that the same techniques can be applied to the quark contributions, which would turn the all-gluon demonstration into a full NNLO description of dijet production.","With the scale band collapsing as shown, jet data become usable in global parton-distribution fits at NNLO rather than only at NLO, and the theory error on $a_s$ extracted from jet cross sections stops being dominated by scale variation.","Because the correction is presented differentially in $p_T$ and rapidity, the size of the NNLO shift can be tracked from central to forward jets, which is where the gluon density is least constrained.","The exclusive dijet invariant mass distribution, the observable most directly tied to PDF fits, now has an NNLO prediction in this channel, with a $k$-factor flat in $m_{jj}$.","The wedge decomposition and local convergence strategy established here is a reusable template for other hadron-collider observables computed with initial-state radiators."],"supporting_citations":[{"why":"Supplies the antenna subtraction method itself, the framework on which the whole calculation rests.","marker":"[27]"},{"why":"Constructs the tree-level double-real antenna subtraction terms for the all-gluon process used in the four-parton channel.","marker":"[39]"},{"why":"Constructs the one-loop single-real subtraction terms used in the three-parton channel and shows the pole structure.","marker":"[40]"},{"why":"Integrates the subtraction terms and combines them with the two-loop matrix elements to demonstrate cancellation of infrared poles; this is the load-bearing prior result the letter relies on.","marker":"[41]"},{"why":"Provides the two-loop virtual matrix elements for the basic $2/rightarrow 2$ process entering the two-parton channel.","marker":"[12]"},{"why":"Provides the one-loop virtual corrections to the single real radiation $2/rightarrow 3$ process.","marker":"[13]"},{"why":"Provides the tree-level $2/rightarrow 4$ matrix elements for the double real radiation channel.","marker":"[14]"},{"why":"Supplies the MSTW08NNLO gluon distribution used for all numerical results, including the LO and NLO evolution.","marker":"[46]"},{"why":"The earlier all-gluon NLO jet study that established the precedent for this approach and that a NLO-limit check would be compared against.","marker":"[44]"}],"fun_headline_variants":["All-gluon NNLO jet calculation trims scale dependence","Gluon-only NNLO corrections lift dijet rates 15-25 percent","First NNLO step for hadron-collider jets, gluons only","All-gluon NNLO jets: method first, data comparison later","NNLO QCD in the all-gluon channel: 25% to 12% k-factors"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing premise is that the antenna subtraction terms built in the authors' earlier papers cancel every infrared pole for these jet-defined observables, which is asserted by citation rather than re-derived or checked against an independent calculation, and no comparison of the generator's NLO limit with the published all-gluon NLO jet results is shown.","fun_headline_variants_meta":{"raw":{"variants":["All-gluon NNLO jet calculation trims scale dependence","Gluon-only NNLO corrections lift dijet rates 15-25 percent","First NNLO step for hadron-collider jets, gluons only","All-gluon NNLO jets: method first, data comparison later","NNLO QCD in the all-gluon channel: 25% to 12% k-factors"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1276,"prompt_tokens":888,"completion_tokens":388,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":297}},"tokens_in":888,"tokens_out":388,"duration_ms":9205,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":297,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T08:04:35.813278+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Run the generator with the NNLO terms switched off and compare the resulting all-gluon jet cross sections against the published NLO all-gluon results of Refs. [42-44]; a mismatch, or a residual uncancelled pole appearing in any of the three channels of Eq. (1), would undercut the claim that the subtraction is complete for jet observables.","supporting_citations":[{"cited_title":"Antenna Subtraction at NNLO","cited_arxiv_id":"hep-ph/0505111","evidence_quote":"Supplies the antenna subtraction method itself, the framework on which the whole calculation rests."},{"cited_title":"Antenna subtraction for gluon scattering at NNLO","cited_arxiv_id":"1003.2824","evidence_quote":"Constructs the tree-level double-real antenna subtraction terms for the all-gluon process used in the four-parton channel."},{"cited_title":"Real-Virtual corrections for gluon scattering at NNLO","cited_arxiv_id":"1112.3613","evidence_quote":"Constructs the one-loop single-real subtraction terms used in the three-parton channel and shows the pole structure."},{"cited_title":"Double Virtual corrections for gluon scattering at NNLO","cited_arxiv_id":"1211.2710","evidence_quote":"Integrates the subtraction terms and combines them with the two-loop matrix elements to demonstrate cancellation of infrared poles; this is the load-bearing prior result the letter relies on."},{"cited_title":"Two-loop QCD corrections to gluon-gluon scattering","cited_arxiv_id":"hep-ph/0102201","evidence_quote":"Provides the two-loop virtual matrix elements for the basic $2/rightarrow 2$ process entering the two-parton channel."},{"cited_title":"One-loop n-point gauge theory amplitudes, unitarity and collinear limits","cited_arxiv_id":"hep-ph/9403226","evidence_quote":"Provides the one-loop virtual corrections to the single real radiation $2/rightarrow 3$ process."},{"cited_title":"Multi-Parton Amplitudes in Gauge Theories","cited_arxiv_id":"hep-th/0509223","evidence_quote":"Provides the tree-level $2/rightarrow 4$ matrix elements for the double real radiation channel."},{"cited_title":"Parton distributions for the LHC","cited_arxiv_id":"0901.0002","evidence_quote":"Supplies the MSTW08NNLO gluon distribution used for all numerical results, including the LO and NLO evolution."},{"cited_title":"Next-to-Leading Order Gluonic Three Jet Production at Hadron Colliders","cited_arxiv_id":"hep-ph/9610433","evidence_quote":"The earlier all-gluon NLO jet study that established the precedent for this approach and that a NLO-limit check would be compared against."}],"review_version":1}