{"id":"ef636bcb-7081-4240-8c4a-717b186f304a","arxiv_id":"2411.14050","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"First extraction of the two-loop universal central-emission vertex in QCD and in N=4 super Yang-Mills from the multi-Regge limit of five-point amplitudes.","lead":"This paper derives the two-loop corrections to the central gluon emission vertex in high-energy QCD scattering by comparing an effective field theory prediction to exact five-point amplitudes. If correct, it supplies a key missing ingredient for next-to-next-to-leading-logarithmic resummations in the Regge limit.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 3-W rapidity-evolution power counting in Section 3.5 is the least externally checked input; an explicit computation is needed to prove it cannot contaminate the extracted U^(2).","rationale":"The reader correctly identifies the rapidity evolution of the 3-W state as the least externally supported assumption in the derivation. The paper's justification is an analogy rather than an explicit computation, and the extracted U^(2) depends on this assumption. I do not think the concern warrants rejection: the standard power counting of the Balitsky/JIMWLK Hamiltonian strongly suggests the effect is one order higher than the two-loop NNLL terms considered. However, because the paper does not demonstrate this explicitly, and because an independent check is cheap and would remove all doubt, the conditional verdict is appropriate. The rest of the central claim is well supported: the MRK expansion is validated against N=4 results, the one-loop vertex agrees with refs. [41,42], and the subleading-colour multi-W terms are independently checked in progress per the acknowledgments. The main soft spot remains the unproven power-counting assertion about 3-W evolution, exactly as the reader stated.","tokens_in":53289,"tokens_out":11577,"duration_ms":126019,"concrete_test":"Compute the O(alpha_s) JIMWLK/BFKL evolution kernel acting on the 3-W state and evaluate its contribution to the {1,3}, {3,1}, and {3,3} NNLL amplitudes of eqs. (3.68) and (3.73). If every such contribution is O(alpha_s^3 L) or higher, the power-counting in Section 3.5 is confirmed. If any piece survives at O(alpha_s^2 L^0) or O(alpha_s^2 L), the extracted U^(2) in eqs. (4.38) and (4.41) is contaminated and the universality interpretation must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3.5 states that the rapidity evolution of the 3-W intermediate states 'starts at three loops' and therefore does not affect the two-loop NNLL extraction, but the justification is only an analogy with the one-loop case in Section 3.2.2. No explicit computation of the leading evolution kernel acting on the 3-W state is presented; eqs. (3.67)-(3.73) evaluate the multi-W amplitudes with free, unevolved W correlators. This matters because U^(2) in eqs. (4.38) and (4.41) is extracted after subtracting precisely these multi-W contributions: any two-loop-order evolution effect on the 3-W state would be absorbed into U^(2) and would change its interpretation as the universal two-loop central-emission vertex. The standard power-counting argument suggests the effect is suppressed: the JIMWLK Hamiltonian is O(alpha_s), so one insertion on a 3-W state contributes O(alpha_s^3 L) relative to the two-loop NNLL amplitude. That argument is plausible but is not actually carried out in the paper, so the universality claim rests on an unproven power-counting step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies 2 -> 3 QCD scattering amplitudes in multi-Regge kinematics at two loops. It expands the known full-colour five-parton amplitudes of ref. [71] in the MRK limit using pentagon-function expansions, defines IR-finite remainders, and compares the result with predictions from the Balitsky/JIMWLK effective theory. At one loop the paper reproduces all signature components from universal impact factors, Regge trajectories and a one-loop central-emission vertex; at two loops it predicts the odd-odd amplitude up to NNLL in terms of known quantities plus multi-W (three-reggeon) exchange contributions and one unknown two-loop vertex U^(2), which is then extracted from the full amplitude. The same procedure is applied to N = 4 sYM, giving the new two-loop result in eq. (4.41).","tokens_in":53475,"tokens_out":13718,"duration_ms":140860,"significance":"If the central interpretation is correct, this is an important step for the Regge limit of QCD: it shows that the apparent non-universality of two-loop MRK amplitudes at NNLL is accounted for by calculable subleading-colour multi-W exchanges, and it provides the first two-loop central-emission vertex in QCD and N = 4 sYM. The paper contains several strong validation elements: direct comparison with the N = 4 sYM amplitudes of ref. [69], agreement with the one-loop Lipatov vertex of refs. [41,42] to O(epsilon^2), independent agreement on multi-W contributions communicated in ref. [105], and reproduction of the subleading-colour partonic-channel dependence. The results are distributed in ancillary files, and the finite remainders are remarkably simple and free of spurious singularities. The extraction of U^(2) is honest in that it is defined as the remainder after subtracting predicted terms; this means the odd-odd NNLL amplitude is not fully ab initio, but the NLL components and the multi-W components are checked independently.","major_comments":[{"comment":"The neglect of rapidity evolution of the three-W intermediate states is load-bearing for the extraction of U^(2) in eqs. (4.38) and (4.41), yet the only justification offered is the sentence 'in analogy with the NLL case at one loop, the evolution starts at three loops so it does not affect our two-loops analysis.' This is not a computation. The power counting should be spelled out: the JIMWLK Hamiltonian is O(alpha_s), so an insertion on a three-W state generates O(alpha_s^3 L) relative to the two-loop three-W contribution, but the statement in Section 3.1 that n -> n+2 transitions are suppressed by 'at least one power of alpha_s' relative to n -> n transitions leaves room for 1W -> 3W transitions at O(alpha_s^2 L), which would appear at two-loop NLL order. Please either prove by explicit action of the evolution Hamiltonian on the three-W state that no O(alpha_s^2 L) or O(alpha_s^2 L^0) evolution effect contributes, or state the precise counting and reconcile it with Section 3.1. If such an effect existed, U^(2) would absorb it and the interpretation of U^(2) as the universal two-loop central-emission vertex would need to be revised.","section":"Section 3.5, eqs. (3.66)-(3.73)"},{"comment":"The extracted finite remainder Uhat^(2) is a scheme-dependent object: it depends on the finite-remainder subtraction in eqs. (4.34)-(4.35), on the reabsorption of the leading-colour multi-W component in eq. (4.15), and on the scales mu and rho. The relation of this Uhat^(2) to the conventional two-loop Lipatov/central-emission vertex is not established; the paper itself notes in Section 5 that a rigorous proof of the relation is lacking. Given the abstract's claim that the paper extracts 'the universal vertex that controls the emission of the central-rapidity gluon', please either provide the two-loop relation to the standard vertex along the lines of eq. (4.42), or explicitly qualify that the extracted object is the scheme-dependent EFT coefficient U defined here, not the standard Lipatov vertex.","section":"Section 4.4, eqs. (4.38)-(4.41) and Section 5"}],"minor_comments":[{"comment":"The independent check against ref. [105] is described only as private correspondence ('We have corresponded with the authors and found agreement'). Since the validation narrative includes this as a key check, the reader cannot assess what was compared. Please document the compared quantities in an appendix or table, or cite a public version where the comparison is shown.","section":"Section 4.5, ref. [105]"},{"comment":"The sentence 'the Regge trajectory tau_g expanded to three loops (though the latter only enters starting from the three-loop level)' is confusing: for the two-loop amplitude the two-loop trajectory is required in eq. (4.9), while the three-loop trajectory enters only in the three-loop amplitude. Please rephrase to distinguish the order of the trajectory from the loop order of the amplitude.","section":"Section 3.5, text after eq. (3.66)"},{"comment":"The phrase 'predict these amplitudes for the first time to NNLL' overstates what is done for the odd-odd two-loop amplitude, since U^(2) is extracted from the same full amplitudes rather than predicted. The body of the paper is careful about this distinction; the abstract and introduction should be qualified accordingly.","section":"Abstract and Section 4.1"}],"recommendation":"major_revision","confidential_remarks":"The technical structure appears sound and the paper is likely correct in its main conclusions, but the revision should make the three-W evolution power counting explicit and calibrate the 'central-emission vertex' claim. These are load-bearing points for the universality interpretation, not cosmetic issues. No grounds for rejection are apparent."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a serious paper and probably the right next step for NNLL BFKL phenomenology. The genuinely new results—the first NNLL prediction of 2 to 3 amplitudes in the Balitsky/JIMWLK framework and the first extraction of the two-loop central-emission vertex in QCD and N=4 sYM—hold up under the checks they report. I believe the extraction is honest: they subtract all objects they can predict (Regge trajectory, impact factors, multi-W coefficients) and the remainder, U^(2), is the vertex. They are clear that this is a matching procedure, not a purely predictive derivation.\n\nThe paper does a lot right. The MRK expansion of the pentagon functions is documented in detail and validated against numerical evaluations. They cross-check the one-loop Lipatov vertex against the existing literature to O(epsilon^2), reproduce the N=4 sYM amplitudes of ref. [69], and agree with independent calculations of the multi-W contributions (ref. [105], communicated). That is a lot of independent confirmation, and the ancillary files make the results usable.\n\nThe softest spot is in Section 3.5. The claim that rapidity evolution of the 3-W intermediate states starts at three loops, hence does not contaminate the two-loop extraction, is justified only by analogy with the one-loop case. No explicit computation of the action of the leading evolution kernel on the 3-W state is given. The standard power counting—the JIMWLK Hamiltonian is O(alpha_s), so one insertion on a 3-W state contributes O(alpha_s^3 L) relative to the NNLL amplitude—makes the statement plausible, but the paper does not actually walk through it. If a two-loop-order evolution effect existed, it would be absorbed into U^(2) and the universality interpretation would need revision. I do not think this is fatal; it is an addressable gap, and the agreement of the extracted vertex with known one-loop structure and with N=4 maximal transcendentality lends some indirect support.\n\nOne minor point: the abstract says the non-universality comes from contributions 'that we can predict.' That is true for the multi-W pieces, but U^(2) itself is extracted, not predicted. The wording is a bit generous, but the body of the paper is clear about the distinction.\n\nWho is this for: anyone working on high-energy factorization, BFKL, small-x, or two-loop amplitudes in the Regge limit. The paper deserves a serious referee and will be cited, at least as the source of the two-loop CEV and the first NNLL MRK comparison. I would send it for peer review without hesitation.","headline":"The first NNLL 2-to-3 MRK prediction and the first two-loop central-emission vertex: a strong, honest extraction whose main soft spot is an unproven power-counting assumption about multi-W evolution.","tokens_in":54076,"tokens_out":2701,"would_cite":true,"duration_ms":27504,"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":"The paper establishes that two-loop five-point QCD amplitudes in the high-energy limit are described by universal building blocks plus predictable subleading-colour multi-reggeon exchanges, and extracts the two-loop central-emission…","keywords":["multi-Regge kinematics","QCD scattering amplitudes","next-to-next-to-leading logarithmic accuracy","reggeons","central-emission vertex","universality","N=4 super Yang-Mills","five-parton amplitudes"],"falsifier":"A direct two-loop calculation of the rapidity evolution (anomalous dimension) of the three-reggeon $WWW$ state would settle the question: if it is non-zero at two loops, the extracted $\\hat U^{(2)}_{\\lambda}$ would absorb those evolution effects and the universality interpretation would need revision. The paper does not provide this calculation.","tokens_in":53058,"feed_emoji":"⚛️","tokens_out":10229,"duration_ms":96248,"temperature":0.7,"pith_summary":"The paper studies $2\\to 3$ quark and gluon scattering in multi-Regge kinematics, where the final particles are strongly ordered in rapidity. Using an effective description in which the exchanged objects are reggeon fields, it predicts these five-point amplitudes to next-to-next-to-leading logarithmic (NNLL) order and compares the prediction with the exact two-loop QCD amplitudes expanded in that limit. It finds that the full-colour results are reproduced by universal objects together with subleading-colour contributions from multiple reggeon exchanges, so the apparent non-universality at NNLL is a predictable effect rather than a breakdown of factorisation. This makes possible the first extraction of the two-loop vertex that controls central-rapidity gluon emission, in both QCD and $N=4$ super Yang-Mills.","feed_headline":"Two-loop gluon-emission vertex extracted in QCD","feed_subtitle":"Five-parton amplitudes in the high-energy limit match universal objects; the rest is predictable subleading-colour physics.","key_machinery":"The central object is the $W$ field, a signature-odd reggeon field in the effective theory used in this paper, whose rapidity evolution is controlled by the gluon Regge trajectory $\\tau_g$ and whose same-rapidity two-point function is free. The argument works by expanding the amplitude in this field: one-$W$ exchange builds the factorised Regge-pole chain, while three-$W$ exchanges generate the multi-reggeon contributions. The load-bearing identity is the NNLL decomposition of eq. (4.16), in which the amplitude splits into a universal single-$W$ chain and a manifestly subleading-colour bracket involving colour operators $T^2_{\\sigma_1\\sigma_2}$; matching this decomposition to the full two-loop amplitude and subtracting infrared poles yields the finite two-loop vertex $\\hat U^{(2)}_{\\lambda}$ that is reported for QCD and $N=4$ super Yang-Mills.","core_discovery":"On the paper's own terms, the central discovery is that the two-loop, signature-odd $2\\to 3$ amplitude in multi-Regge kinematics can be written exactly as a factorised single-reggeon chain, built from Regge trajectories, impact factors, and the central-emission vertex, plus explicit $\\pi^2$ contributions from multi-$W$ (three-reggeon) exchanges, and that this reproduces the infrared-subtracted full-colour QCD amplitude. The single-reggeon chain contains a universal two-loop vertex $\\hat U^{(2)}_{\\lambda}$, given in eqs. (4.38) and (4.41), that was previously unknown; once it is extracted, every other term in the NNLL prediction is either a known quantity or a predicted subleading-colour multi-$W$ contribution. The authors therefore claim that the apparent non-universality seen at NNLL in the full amplitudes is fully accounted for by terms they can predict, and that the universal vertex is the last missing ingredient for NNLL signature-odd multi-Regge amplitudes at any multiplicity.","pith_inferences":["Beyond the paper's five-point extraction, the same two-loop vertex should appear in the multi-Regge limit of six-point QCD amplitudes once they are computed, giving an independent test of its universality.","Beyond the paper's two-loop order, the deferred rapidity evolution of the three-reggeon state is the most plausible place for the universality picture to be revised; if that evolution begins at two loops, the extracted vertex would absorb those effects.","Beyond the paper's explicit results, the simplicity of the $N=4$ vertex suggests that it may obey a differential equation in the transverse variables $z$ and $\\bar z$, which would provide a separate consistency check."],"forward_implications":["The extracted two-loop central-emission vertex completes the set of universal ingredients needed for NNLL signature-odd multi-Regge predictions, so future higher-multiplicity two-loop amplitudes can be checked against it.","The subleading-colour multi-$W$ terms are shown to account for all flavour dependence of the apparent non-universality at NNLL, meaning no new unknown process-dependent function is required at this order.","The $N=4$ super Yang-Mills two-loop vertex equals the leading-transcendental part of the QCD vertex, so the maximal-transcendentality relation holds for this building block as well.","Within the framework used here, combining the extracted single-gluon vertex with the known two-gluon emission vertex should make fully universal NNLL predictions for $n$-point multi-Regge amplitudes possible in principle."],"supporting_citations":[{"why":"Provides the effective framework from which the factorised MRK amplitude and its $W$-field expansion are taken.","marker":"[44]"},{"why":"Supplies the full-colour two-loop five-parton QCD amplitudes whose MRK limit is compared against the predictions.","marker":"[71]"},{"why":"Supplies two-loop quark-channel five-parton amplitudes used for full-colour validation.","marker":"[72]"},{"why":"Supplies two-loop gluon-channel five-parton amplitudes used for full-colour validation.","marker":"[73]"},{"why":"Provides the two-loop five-point amplitudes in $N=4$ super Yang-Mills in the multi-Regge limit, used for the $N=4$ extraction and the leading-transcendental comparison.","marker":"[69]"},{"why":"Defines the separation between universal leading-colour and subleading-colour multi-reggeon contributions and supplies two-loop impact-factor inputs.","marker":"[31]"},{"why":"Provides the three-loop gluon Regge trajectory used as a known ingredient at NNLL accuracy.","marker":"[30]"},{"why":"Provides the one-loop central-emission vertex through higher orders in the dimensional regulator, used to validate the one-loop vertex extraction.","marker":"[42]"}],"fun_headline_variants":["Universal two-loop vertex extracted from five-point QCD","Apparent QCD non-universality pinned to predictable terms","Single universal vertex explains five-parton QCD at NNLL","Missing two-loop vertex found, QCD amplitudes unified"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the rapidity evolution of the three-reggeon (multi-$W$) intermediate state starts only at three loops, so it does not affect the two-loop vertex extraction; the paper justifies this by analogy with the one-loop case rather than by an explicit computation.","fun_headline_variants_meta":{"raw":{"variants":["Universal two-loop vertex extracted from five-point QCD","Apparent QCD non-universality pinned to predictable terms","Single universal vertex explains five-parton QCD at NNLL","Missing two-loop vertex found, QCD amplitudes unified"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000535,"raw_usage":{"total_tokens":2560,"prompt_tokens":924,"completion_tokens":1636,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":540,"completion_tokens_details":{"reasoning_tokens":1569}},"tokens_in":540,"tokens_out":1636,"duration_ms":13556,"temperature":1.0,"reasoning_tokens":1569,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:35:30.239119+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct two-loop calculation of the rapidity evolution (anomalous dimension) of the three-reggeon $WWW$ state would settle the question: if it is non-zero at two loops, the extracted $\\hat U^{(2)}_{\\lambda}$ would absorb those evolution effects and the universality interpretation would need revision. The paper does not provide this calculation.","supporting_citations":[],"review_version":1}