{"id":"691306ae-16b9-4e6b-8139-9df1f82dc13c","arxiv_id":"2412.06519","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"First analytic full-color two-loop five-particle amplitudes for associated Higgs-plus-bottom-quark-pair production at the LHC, with public C++ implementation.","lead":"This paper computes the complete two-loop quantum corrections for producing a Higgs boson together with a bottom-quark pair at the LHC, including all color structures. The analytic results come with a public C++ library, opening the way to more precise predictions for a key Higgs coupling measurement.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"All two-loop validation checks share the same IBP reduction and master-integral expansion, so a systematic error in that shared pipeline would survive every listed check; an independent numerical two-loop evaluation is the missing test.","rationale":"The reader's weakest assumption is exactly the shared-validation-pipeline problem: the two-loop result is checked internally, but the checks reuse the same IBP reduction, master-integral basis, pentagon-function expansion, and finite-field reconstruction. I agree that this is the load-bearing weakness. It is a real epistemic gap, because a constant finite error respecting gauge invariance and factorisation would pass all listed tests. However, the paper is transparent about the validation, provides public analytic expressions and a C++ library, and the checks performed are consistent with current standards for multi-loop amplitude computations. The leading-colour part also has a prior published result, though that does not cover the subleading-colour structures. Given these circumstances, I do not regard the missing independent numerical benchmark as a reason to reject or to demand reclassification; it is a caveat and a natural next validation step. The verdict ACCEPT therefore remains appropriate, and the concern should be recorded as a recommendation for an independent cross-check rather than as a demonstrated flaw.","tokens_in":33639,"tokens_out":8132,"duration_ms":100939,"concrete_test":"At the benchmark phase-space point of Eqs. (4.3)-(4.4), evaluate the two-loop finite remainder for the most complex subleading-colour partial amplitude, e.g. A(2),1_34 or A(2),Nc_delta in Eq. (2.15), with an independent chain: generate diagrams with QGraf, perform IBP reduction with Kira or FIRE instead of NeatIBP, and evaluate the master integrals numerically with AMFlow or pySecDec at the same point. Compare the resulting finite remainder against the public C++ library to at least 10^-6 relative accuracy; agreement would close the loophole, while disagreement would localise a systematic error in the shared reduction or reconstruction pipeline.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3.3 lists several checks, but none independently exercises the two-loop reduction. The direct-helicity comparison is a genuine alternative construction of the loop numerators, yet both sides are reduced with the same NeatIBP systems (Eq. (3.12)), mapped through the same LiteRed master-integral relations, and reconstructed through the same finite-field graphs. The Ward-identity check constrains gauge invariance at the master-integral-coefficient level and cannot detect a reduction error that respects gauge invariance. Pole cancellation and the mu-dependence test (Eq. (3.24)) constrain the singular and scale-dependent parts, which are built from the same finite remainders and lower-loop inputs; they say nothing about the finite part at mu = 1. The OpenLoops comparison is tree-level and one-loop only. Thus the entire two-loop validation is insensitive to a systematic error in, for example, the DPmz/DPzz IBP systems of Table 1 or the permutation rules connecting ordered integral families. This is the weakest link in the claim that the analytic full-colour result is correct.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the first analytic full-colour two-loop five-particle scattering amplitudes for pp → b bbar H in the five-flavour scheme, with massless bottom quarks and a finite bottom-Yukawa coupling. The amplitudes are obtained through finite-field reconstruction, decomposed into colour and helicity structures, and expressed in terms of one-mass pentagon functions. The authors provide a public C++ library and Mathematica ancillary files for the finite remainders and pole terms, and they report benchmark values of the hard functions together with a numerical stability analysis. The paper claims to be the first analytic full-colour two-loop five-point amplitude with an external mass, extending earlier leading-colour results.","tokens_in":33846,"tokens_out":6218,"duration_ms":63805,"significance":"If the result is correct, this is a major technical achievement and an important step toward NNLO QCD phenomenology for b-bbar-H production, as well as for massification-based approximations of other processes. The manuscript includes several strengths: analytic expressions for all colour structures, a publicly available C++ implementation, detailed documentation of the finite-field reconstruction strategy, multiple internal validation checks, and a stability study in a realistic phase-space. The result is not fitted to data and appears free of circularity. The main weakness is that the two-loop validation lacks a fully independent external numerical benchmark, as all checks share the same IBP reduction and master-integral expansion.","major_comments":[{"comment":"All two-loop validation checks listed in Section 3.3 share the same IBP reduction (Eq. (3.12)), the same LiteRed master-integral relations, the same one-mass pentagon-function expansion, and the same finite-field reconstruction. A systematic error in, for example, the DPmz/DPzz IBP systems of Table 1 or in the permutation rules connecting ordered integral families would therefore survive every listed check. The Ward-identity test constrains gauge invariance at the master-integral-coefficient level; pole cancellation and the mu-scaling test (Eq. (3.24)) constrain only the singular and scale-dependent parts; the OpenLoops comparison is tree-level and one-loop only. The direct-helicity comparison is an alternative construction of the loop numerators, but both sides are reduced with the same NeatIBP systems and reconstructed with the same finite-field graphs. I recommend that the authors provide an independent numerical two-loop evaluation for at least one phase-space point, for example using a different reduction code (e.g., Kira or FIRE) or a numerical unitarity/sector-decomposition approach, or else explicitly discuss this limitation and justify why the existing internal consistency checks are sufficient for the central claim of correctness.","section":"Section 3.3"}],"minor_comments":[{"comment":"There is a typo: 'to go from Eq. (3.16) to Eq. (3.16)' should read 'to go from Eq. (3.11) to Eq. (3.16)'.","section":"Section 3.2, after Eq. (3.17)"},{"comment":"The phrase 'only2 − 3 permille of the points' should be written as 'only 2–3 per mille (0.2%–0.3%) of the points' for clarity.","section":"Section 4.2"},{"comment":"The word 'Mathematicascripts' appears without a space; it should be 'Mathematica scripts'.","section":"Appendix A"},{"comment":"In Eq. (2.22), the colour decomposition for the b-bbar-b-bbar channel reuses the symbols A1 and A2 from the q-qbar case while introducing A3 and A4; a sentence clarifying that these are the same partial amplitudes and that the relations in Eq. (2.23) implement the required permutations would improve readability.","section":"Section 2.3"},{"comment":"It would be helpful to state explicitly that the benchmark values in Table 3 are not cross-checked with an external independent implementation at two loops, consistent with the validation limitation noted in Section 3.3.","section":"Section 4.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is a significant technical achievement and the internal validation is substantial, but the absence of an independent two-loop numerical cross-check is the main obstacle to full confidence. If the authors can add such a check or convincingly argue that the shared-pipeline risks are mitigated, I would support acceptance. Otherwise, the manuscript should be revised to include the requested validation or to explicitly frame the remaining uncertainty in the correctness claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is the real thing: the first analytic full-colour two-loop five-particle amplitude for a process with one external mass (pp -> b bbar H in the five-flavour scheme). The leading-colour version existed [42], and W gamma gamma got full colour numerically [91]; this closes the analytic gap. The authors derive finite remainders for all colour structures, write them in terms of one-mass pentagon functions, provide public C++ code, and give a stability analysis over 100k phase-space points. That is a major advance.\n\nWhat they do well: the computation is documented carefully - colour decompositions, tensor projectors, IBP reduction with NeatIBP, finite-field reconstruction with a long list of optimisations. Validation is extensive by field standards: direct-helicity cross-check (an alternative numerator construction), Ward identity, pole cancellation, mu-dependence, OpenLoops comparison at tree and one loop, plus a cross-check of the C++ implementation against Mathematica. The ancillary files and code are available, which counts for a lot.\n\nThe soft spot is exactly the one you suspect: all two-loop checks share the same IBP reduction, the same master-integral basis, the same pentagon-function expansion, and the same finite-field reconstruction. A systematic error in, say, the DPmz/DPzz systems or the permutation rules would survive every listed check. The direct-helicity comparison only re-derives the numerator; both sides then go through the same reduction. Ward identity and pole cancellation constrain the singular parts, not the finite part at mu = 1. OpenLoops only goes to one loop. So there is no fully independent external numerical benchmark at two loops. This is a real gap, and I would rate soundness a notch below the top-tier papers for that reason. But it is also the standard validation level in this field; nothing in the paper makes me think the result is wrong.\n\nThe circularity concern in the reader's report is a non-issue: reusing one's own IBP infrastructure and the earlier leading-colour result is normal engineering, not circular reasoning. No parameters are fitted to data.\n\nWho is this for? Anyone doing NNLO QCD for H + bottom jets, and amplitude practitioners. It deserves a serious referee. My recommendation: send to a strong referee who knows the pipeline, ask them to scrutinise the IBP permutation logic and, if feasible, push for an independent numerical two-loop check at a few phase-space points. Accept after revision - this is a milestone paper even if the validation could be one notch stronger.","headline":"First full-colour two-loop five-point amplitude with a massive external leg; well validated internally but no independent two-loop numerical check - send to a serious referee.","tokens_in":34374,"tokens_out":2463,"would_cite":true,"duration_ms":27131,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["12.38.Bx","13.85.-t","14.80.Bn"],"model":"deepseek-v4-flash","headline":"For the first time, a full-colour two-loop five-particle amplitude with an external mass is computed analytically, for Higgs production with a bottom-quark pair.","keywords":["Higgs boson production","bottom-quark pair","two-loop amplitudes","full colour","five-point scattering","helicity amplitudes","finite-field reconstruction","NNLO QCD"],"falsifier":"Compute the two-loop finite remainder at the benchmark phase-space point (or any other point) using an independent method that does not share the same integral-reduction, master-integral, and pentagon-function machinery, and compare the result: any disagreement beyond the stated numerical precision would show that the analytic amplitude is wrong.","tokens_in":33473,"feed_emoji":"⚛️","tokens_out":5657,"duration_ms":58376,"temperature":0.7,"pith_summary":"This paper reports the first analytic, full-colour two-loop scattering amplitude for a five-particle process with one massive external leg: Higgs boson production in association with a bottom-quark pair at the LHC. The computation is done in the five-flavour scheme, where the bottom quark is massless but its Yukawa coupling to the Higgs stays finite, and all colour structures are retained. If correct, the result supplies the double-virtual ingredient needed for NNLO QCD predictions for $pp \\to b\\bar{b}H$, and it opens the way to similar external-mass processes. The authors provide the finite remainders as a public C++ library and demonstrate that evaluating them is fast and numerically stable enough for phenomenological studies.","feed_headline":"First full-colour two-loop amplitudes for b-quark pair plus Higgs","feed_subtitle":"All colour structures now included for pp to b bbar H, making NNLO QCD predictions possible.","key_machinery":"The calculation is carried by decomposing the amplitudes into partial colour structures, reducing all Feynman integrals to a minimal pure basis of master integrals, and expanding those in one-mass pentagon functions: special functions describing five-point integrals with one massive external leg, here the Higgs boson. The rational coefficients multiplying these pentagon functions are reconstructed from numerical evaluations over finite fields, using momentum-twistor variables to rationalise the kinematics. The colour decomposition, the integration-by-parts reduction to master integrals, the pentagon-function expansion, and the finite-field reconstruction together form the machinery that makes the full-colour analytic result tractable.","core_discovery":"The authors establish that the two-loop helicity amplitudes for the partonic processes contributing to $pp \\to b\\bar{b}H$ in the five-flavour scheme can be expressed analytically in terms of one-mass pentagon functions with rational coefficients reconstructed over finite fields, retaining the complete colour structure. This is the first time an analytic full-colour two-loop five-point amplitude with an external mass has been obtained. The finite remainders are free of UV and IR poles by construction; the paper provides their analytic expressions, the pole terms, and benchmark hard functions for all partonic channels. The authors also report that the complete two-loop contribution is roughly ten percent of the leading-order cross section, with subleading-colour terms contributing about two percent.","pith_inferences":["If the central claim is right, the main bottleneck for NNLO $b\\bar{b}H$ phenomenology shifts from the double-virtual amplitudes to the real-emission and subtraction pieces, and differential predictions with full colour should now be within reach.","The observed cancellations of the square-root letter $\\sqrt{\\Delta_5}$ in the finite remainders and of the non-planar letters $\\sqrt{\\Sigma_5^{(i)}}$ in the bare amplitudes suggest universal analytic structures for one-mass five-point amplitudes, which could be proven from factorisation and analyticity arguments.","A fully independent numerical evaluation of one benchmark phase-space point, using a method that does not share the same integral-reduction and pentagon-function pipeline, would provide a stronger external check than the internal consistency tests reported in the paper.","The same colour-decomposition and finite-field reconstruction strategy could be applied to other external-mass five-point processes, such as $W b\\bar{b}$ or $Z b\\bar{b}$ production, where full-colour two-loop results are not yet available analytically."],"forward_implications":["The analytic finite remainders provide the double-virtual ingredient needed for NNLO QCD predictions for $pp \\to b\\bar{b}H$ with two bottom-tagged jets in the five-flavour scheme.","Full colour control allows the subleading-colour contributions to be quantified, which the paper estimates at about two percent of the leading-order cross section.","The public C++ implementation, with average evaluation times near forty seconds per phase-space point and a precision-rescue strategy, is ready for use in Monte Carlo phenomenology.","Through the massification procedure that restores leading bottom-mass effects, the authors argue the results can also approximate four-flavour-scheme predictions and high-energy $t\\bar{t}H$ amplitudes."],"supporting_citations":[{"why":"Supplies the previous leading-colour two-loop amplitude for $b\\bar{b}H$ production that this work extends to full colour.","marker":"[42]"},{"why":"Provides the full-colour $W\\gamma\\gamma$ amplitude framework, including optimised integration-by-parts relations that are reused here.","marker":"[91]"},{"why":"Defines the complete set of two-loop five-point one-mass master integrals and pentagon functions used to express the amplitude.","marker":"[53]"},{"why":"Gives the planar two-loop five-point one-mass master integrals that form part of the reduction basis.","marker":"[49]"},{"why":"Gives the non-planar hexa-box integrals for one-mass five-point processes needed for the full colour result.","marker":"[51]"},{"why":"Generates the optimised integration-by-parts relations used to reduce the large set of scalar Feynman integrals to master integrals.","marker":"[132]"},{"why":"Provides the finite-field functional-reconstruction framework used to obtain the analytic rational coefficients from numerical samples.","marker":"[60]"},{"why":"Supplies the helicity-construction, permutation, and numerical-rescue strategies adopted for evaluating the amplitude stably.","marker":"[76]"},{"why":"Used to cross-check the tree-level and one-loop hard functions against independent numerical results.","marker":"[141]"}],"fun_headline_variants":["First analytic full-colour two-loop amplitudes for pp to bbH","Analytic full-colour two-loop amplitudes for bbH production at LHC","Complete colour structure at two loops for Higgs plus bottom pair","First analytic full-colour two-loop five-point amplitudes with massive leg"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the internal consistency checks, which all share the same reduction and pentagon-function pipeline, are sufficient to certify the two-loop result; an independent two-loop numerical evaluation is not performed.","fun_headline_variants_meta":{"raw":{"variants":["First analytic full-colour two-loop amplitudes for pp to bbH","Analytic full-colour two-loop amplitudes for bbH production at LHC","Complete colour structure at two loops for Higgs plus bottom pair","First analytic full-colour two-loop five-point amplitudes with massive leg"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001478,"raw_usage":{"total_tokens":5857,"prompt_tokens":784,"completion_tokens":5073,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":400,"completion_tokens_details":{"reasoning_tokens":4998}},"tokens_in":400,"tokens_out":5073,"duration_ms":32427,"temperature":1.0,"reasoning_tokens":4998,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T19:33:51.084625+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the two-loop finite remainder at the benchmark phase-space point (or any other point) using an independent method that does not share the same integral-reduction, master-integral, and pentagon-function machinery, and compare the result: any disagreement beyond the stated numerical precision would show that the analytic amplitude is wrong.","supporting_citations":[],"review_version":1}