{"id":"6adc450d-dc5e-4d34-a9f3-381f42384e4a","arxiv_id":"2501.01847","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Planar two-loop six-particle massless master integrals are solved analytically up to weight four as Chen iterated integrals, with a complete 245-letter alphabet and a validated numerical implementation.","lead":"This paper derives the complete analytic solution space for all planar two-loop six-particle Feynman integrals, expressing them as iterated integrals up to transcendental weight four. It removes a major bottleneck toward next-to-next-to-leading-order predictions for 2-to-4 particle collider processes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Completeness of the function space depends on an unproven reduction-coefficient claim that is broader than the tested set; the finite-epsilon reduction argument in §4 is not fully closing that gap.","rationale":"The reader's weakest assumption is almost exactly the concern I identified: the finite-epsilon reduction-coefficient claim, proved only on a maximal-cut complete set and a selection of off-cut integrals, underlies the paper's completeness claim. I agree this is load-bearing. My proposed test sharpens the reader's concern by isolating the specific place where the argument could fail: the six-dimensional double-pentagon master integral's epsilon expansion. If that integral has a weight-five O(epsilon^1) piece and an O(1/epsilon) coefficient can multiply it, the finite part could receive unprotected higher-weight terms, breaking the 945-symbol counting. The reader's conditional verdict remains appropriate: the paper gives strong partial evidence (40-digit agreement with AMFlow, complete DE for all other sectors, symbol counts) but the completeness of the function space rests on an unproven reduction-coefficient statement. A direct verification of precisely that statement would settle it; if it passes, the verdict can stand or be upgraded. I therefore recommend UNCHANGED, meaning the reader's CONDITIONAL verdict should be kept. Honest non-finding is not appropriate here: the gap is real, specific, and addressable, so the conditional verdict is the right outcome and my concern is a refinement of, not a departure from, the reader's reasoning.","tokens_in":9772,"tokens_out":2118,"duration_ms":20184,"concrete_test":"Take the double-pentagon top-sector master integral I^DP_4 of reference [38] and expand it in epsilon to one more order analytically (from its six-dimensional differential equation) or numerically with AMFlow at the four points x1...x4. Check whether its O(epsilon^1) coefficient has a nonzero weight-five part. In parallel, run full IBP reduction without maximal cut for a substantially larger set of numerator integrals (including higher-rank loop-momentum and Gram-determinant numerators) in the DP and HB families, and verify that the reduction coefficient of I^DP_4 is exactly O(1/epsilon) with no finite or O(epsilon) part. If I^DP_4 = O(epsilon^2) with the epsilon^2 piece already weight six and no weight-five epsilon^1 piece, the §4 extrapolation is safe; if not, the function space may be missing symbols.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the 167-letter/945-symbol function space is complete for evaluating any planar two-loop six-particle amplitude up to the finite part. The load-bearing step is the claim that reduction coefficients for any Yang-Mills numerator contract to finite values as epsilon→0, except for the six-dimensional double-pentagon master integral, whose coefficients may be O(1/epsilon) but whose integral is O(epsilon^2). This is asserted in §4, but verified only on (i) a full set up to rank six on the maximal cut and (ii) a selection of typical integrals without cut. That is not a proof for arbitrary numerators. The transcendental-weight counting argument is also not fully airtight: discarding I^DP_4 as O(epsilon^2) controls its leading epsilon order, but if this UT integral has a weight-five contribution at O(epsilon^1), an O(1/epsilon) reduction coefficient would produce a weight-four term in the finite part that is not covered by the claimed 945 symbols. The paper's checks do not rule out such a contribution for unlisted numerators. The claim is plausible and supported by partial checks, but as presented it remains an extrapolation rather than a derivation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives canonical differential equations for all planar two-loop massless six-particle master integrals, determines analytically the boundary values, and expresses the solutions as Chen iterated integrals. For the double-pentagon top sector it argues that only integrals already known from N=4 sYM contribute up to weight four, and it identifies the complete function space relevant for four-dimensional Yang-Mills amplitudes up to that weight: 167 active alphabet letters and 945 independent symbols (Table I). The claims are supported by finite-field verification of the epsilon-factorized form of the differential equations, by analytic solutions whose one-fold integral representation is checked against AMFlow at four Euclidean points to 40 digits, and by cross-checks against Landau and Baikov analyses.","tokens_in":10016,"tokens_out":5426,"duration_ms":50910,"significance":"If the completeness claim holds, this is a substantial result: it removes the master-integral bottleneck for planar two-loop 2-to-4 massless scattering and provides a concrete symbol alphabet, suitable for bootstrap and for studies of the analytic structure. The paper ships machine-checkable ancillary data (differential equation matrices, boundary values, and symbol lists) and validates the numerical evaluation against an independent method, which are clear strengths. The central caveat is that the completeness argument relies on a regularity property of IBP reduction coefficients that is checked on a finite set of integrands but not proven for arbitrary Yang-Mills numerators; since the symbol count in Table I is complete only under that assumption, the overall claim is plausible but not yet fully established.","major_comments":[{"comment":"The paragraph 'We argue that it is sufficient to expand the basis integrals up to weight four' bases the completeness claim on the assertion that reduction coefficients of any integral appearing in Yang-Mills amplitudes are finite as epsilon tends to zero, except for the six-dimensional double-pentagon integral I^DP_4. The stated support is a complete set of integrals up to rank six on the maximal cut and a selection without cut. This is a finite set of checks, not a proof for arbitrary numerators. Because the 945-symbol space in Table I is complete only if this property holds for all integrals that can appear, please either supply a general argument (e.g., from unitarity or from the structure of the IBP reduction on maximal cuts) or explicitly state that the completeness claim is conditional on this unproven reduction regularity.","section":"§4"},{"comment":"The statement that I^DP_4 'starts contributing at weight six only, i.e. at order epsilon^2' is load-bearing: an O(1/epsilon) reduction coefficient times this integral contributes to the finite part only if the integral has an O(epsilon^1) term. The paper does not derive this epsilon-expansion property; it is asserted. As written, the argument does not rule out an unlisted numerator whose reduction has an O(1/epsilon) coefficient while I^DP_4 has a weight-five O(epsilon^1) term, which would produce a weight-four finite contribution not covered by the 945 symbols. Please provide a derivation or a direct check of the epsilon expansion of I^DP_4 to the required order.","section":"§4"},{"comment":"The differential equation matrix is stated to be incomplete for one six-dimensional double-pentagon entry, and the missing matrix elements 'may or may not be of the form of eq. (8)'. The paper argues this is harmless because the top-sector integrals are either evanescent or known from N=4 sYM. That argument is only as strong as the reduction-coefficient claim of §4; if that claim fails, the missing entry could in principle introduce new functions beyond the 167-letter alphabet even in the finite part. Please make this logical dependence explicit and, if possible, characterize or reconstruct the missing entry to remove the residual uncertainty.","section":"§3"}],"minor_comments":[{"comment":"In the introduction, 'startingfromresults' should read 'starting from results'.","section":"Introduction"},{"comment":"The sentence 'This is a breakthrough in the field of analytic Feynman integral computation provides the relevant information...' is missing a conjunction or a full stop; please rephrase, e.g. '...computation and provides the relevant information...'.","section":"Conclusion"},{"comment":"The abbreviations PB and DB are used without definition, e.g. in 'the top sector integrals of HB, PB and DB'; please define them at first use.","section":"§5"},{"comment":"The phrase 'In comparison with our results, we find that this method fails to identify 18 distinct components...' is unclear because the antecedent of 'this method' is not stated; please specify that it refers to the computational algebraic geometry approach of Refs. [18,19].","section":"§5"},{"comment":"The sentence following eq. (11), 'where the dot indicates a power of a13 = 2', is ambiguous; please clarify whether the dot denotes raising the propagator with index a13 to the second power.","section":"§4"}],"recommendation":"major_revision","confidential_remarks":"The paper overlaps with the recent preprint [62]; the authors note this but do not discuss the extent of the overlap. The main risk for the journal is the unproven reduction-coefficient regularity in §4, which is the load-bearing step for the completeness claim. Given that the rest of the computation is carefully cross-checked, I believe the issue is fixable, but it requires either a proof or a substantial strengthening of the evidence, and the authors should be asked to make the conditional nature of the completeness claim explicit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this is the real thing: the first complete analytic solution space for planar two-loop six-particle massless master integrals up to weight four, and the top sector of the double pentagon turns out to be either evanescent or known from N=4 sYM. That is a striking and useful outcome, not a corner cut.\n\nWhat the paper actually does: it builds the full uniform-transcendental basis for the double-pentagon (267 masters) and hexagon-box (202 masters) families, derives canonical differential equations, fixes analytic boundary values in the Euclidean region, and expresses all solutions as Chen iterated integrals. The alphabet is 245 letters, 167 active through weight four, with 945 independent symbols at weight four (Table I). The numerical validation is strong: 40-digit agreement with AMFlow at four Euclidean points, plus cross-checks against Landau-analysis and Baikov approaches. The auxiliary files appear complete, including symbol lists, weight-two solutions, and a proof-of-concept evaluation code. Citation practice is fine here—the self-citations point to previously published building blocks (five-point one-mass results, the authors' earlier hexagon-box/DP work), not to the target content.\n\nThe one soft spot is the completeness claim in the title and abstract: that the computed function space suffices for any planar two-loop six-particle amplitude up to the finite part. That hinges on the statement that reduction coefficients for Yang-Mills numerators are finite as epsilon goes to zero, except for the 6D double-pentagon integral whose poles are harmless because the integral is O(epsilon^2). This is verified on a complete set of integrals up to rank six on the maximal cut and on a typical selection without cut, but it is not proven for all possible numerators. The stress-test worry about a weight-five contribution at O(epsilon) is directly answered by the paper's explicit statement that I^DP_4 starts at weight six, i.e., O(epsilon^2)—so that specific hole is smaller than it looked. What remains is the reduction-coefficient finiteness for untested numerators, and the missing DE matrix entry for that one 6D double-pentagon integral. The authors are honest about this: they say \"argue\" and \"support this claim,\" not \"prove.\" The gap is real but narrow; it does not shake the hard computational results.\n\nIf I were the editor I would send this to a serious referee without hesitation. The referee should ask for a sharper statement of the unproven reduction claim—ideally a proof or a precise conjecture, and a discussion of whether a full IBP reduction for arbitrary numerators is feasible—but the paper should be published essentially as is. It will be the benchmark for 2-to-4 NNLO calculations and for formal function-algebra studies. Anyone working on six-particle amplitudes or on Landau/cluster methods should read it.\n\nRecommendation: peer review, with targeted but not hostile requests.","headline":"Major computational milestone—first complete analytic solution space for planar two-loop six-particle integrals up to weight four—with one honest gap in the completeness argument that should not block publication.","tokens_in":10523,"tokens_out":2003,"would_cite":true,"duration_ms":20647,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81Q30","81T18"],"pacs":["11.15.-q","11.15.Bt"],"model":"deepseek-v4-flash","headline":"This paper derives the complete system of canonical differential equations for all planar two-loop massless six-particle master integrals, fixes the analytic boundary conditions, and shows this is sufficient to evaluate any…","keywords":["planar two-loop six-particle amplitudes","canonical differential equations","master integrals","Chen iterated integrals","symbol alphabet","uniform transcendental weight","momentum twistors","NNLO QCD"],"falsifier":"A single legitimate Yang-Mills two-loop six-particle numerator whose integration-by-parts reduction onto the uniform-weight basis produces a $1/\\epsilon$ coefficient multiplying an integral that starts at order $\\epsilon^0$ would break the finite-part argument; likewise, a complete amplitude evaluation that requires any alphabet letter or symbol outside the 167 letters and 945 symbols at weight four would falsify the claimed completeness.","tokens_in":9550,"feed_emoji":"⚛️","tokens_out":8526,"duration_ms":76385,"temperature":0.7,"pith_summary":"This paper derives the complete system of canonical differential equations for all planar two-loop massless six-particle Feynman integrals, fixes the analytic boundary values, and writes every solution as a Chen iterated integral. The authors argue that this information is sufficient to evaluate any four-dimensional Yang-Mills scattering amplitude up to its finite part, because two-loop amplitudes have transcendental weight at most four and the uniform-weight master integrals cannot meet reduction coefficients with poles in $\\epsilon$ without producing forbidden higher-weight contributions. They identify the complete function space up to weight four: 167 active alphabet letters and 945 independent symbols, including products of one-loop integrals. The results are validated numerically against direct Feynman-integral evaluation in the Euclidean region, removing the bottleneck of integral evaluation for planar two-to-four processes.","feed_headline":"Complete function space derived for two-loop six-particle amplitudes","feed_subtitle":"All planar massless master integrals become Chen iterated integrals, clearing the way for NNLO 2-to-4 predictions.","key_machinery":"The load-bearing mechanism is the canonical differential equation system for uniform-transcendental-weight (UT) master integrals, $d\\vec I = \\epsilon\\, d\\tilde A\\, \\vec I$, constructed in momentum-twistor variables. The connection matrix is expanded as $\\tilde A = \\sum_j c_j \\log \\alpha_j$, where the alphabet letters $\\alpha_j$ are algebraic functions of the kinematics; the alphabet is seeded by known five-point one-mass letters and extended with new algebraic letters generated by a leading-singularity method. Solving the system as Chen iterated integrals requires analytic boundary values, which are fixed by imposing regularity throughout the Euclidean region. The argument that this suffices for amplitudes has three parts: two-loop amplitudes in four dimensions have weight at most four; a UT integral of the same weight as the amplitude cannot meet an $\\mathcal{O}(1/\\epsilon)$ reduction coefficient without producing a forbidden higher-weight finite term; and the only exceptions are the six-dimensional double-pentagon integral (order $\\epsilon^2$) and evanescent integrals, which vanish at weight four.","core_discovery":"The central claim is that the full set of planar two-loop massless six-particle master integrals — 267 double-pentagon and 202 hexagon-box integrals — satisfies a canonical, $\\epsilon$-factorized differential equation $d\\vec I = \\epsilon\\, d\\tilde A\\, \\vec I$ whose connection matrix is a sum of rational coefficients times logarithms of alphabet letters. Boundary values are fixed analytically at a reference point by regularity in the Euclidean region. The most complicated double-pentagon top sector contributes nothing new up to weight four: its five uniform-weight integrals are either evanescent in four dimensions or reduce, up to $\\mathcal{O}(\\epsilon)$, to the known dual-conformal double-pentagon integrals $\\Omega_{\\mathrm{even}}$ and $\\tilde\\Omega_{\\mathrm{odd}}$ from maximally supersymmetric Yang-Mills theory. The solution space up to weight four therefore consists of 167 letters (11 of them genuine six-particle letters) and 945 independent symbols, of which 45 are genuine two-loop six-point symbols. All master integrals are provided as iterated integrals that can be evaluated numerically.","pith_inferences":["Editorial inference: if the $\\epsilon$-finiteness of reduction coefficients holds for arbitrary numerators beyond rank six and the maximal cut, the completeness claim extends verbatim to amplitudes with higher-dimensional operators or effective-field-theory numerators, not just standard Yang-Mills ones.","Editorial inference: the 167-letter alphabet and 945-symbol tables are likely to reappear as building blocks in neighboring computations (three-loop five-point, one off-shell leg, or non-planar two-loop six-point), so the ancillary files could serve as a shared dictionary.","Editorial inference: the single letter $\\alpha_{100}$, missed by the Baikov-based method, is a sharp test; a physical amplitude that requires $\\alpha_{100}$ at weight two or three would confirm the exception and indicate a systematic gap in Baikov-based alphabet prediction.","Editorial inference: extending this program beyond weight four would require new integration kernels for the double-pentagon top sector; their number could be probed numerically at weight five via high-precision direct evaluation at one kinematic point and then compared against the current 167-letter alphabet."],"forward_implications":["Any planar two-loop massless six-particle amplitude in four dimensions can be evaluated up to the finite part using the provided master integrals, boundary values, and iterated-integral representation.","The complete alphabet (167 letters) and symbol space (945 independent symbols up to weight four) provide the input for bootstrap methods and for studies of analytic structure such as Steinmann relations and factorization.","The removal of the Feynman-integral bottleneck enables NNLO predictions for $2\\to 4$ massless QCD processes.","The analytic solutions allow systematic study of physical limits, including multi-Regge, collinear, and double-parton-scattering limits.","The comparison with Landau-singularity methods identifies 18 singular-locus components those methods miss, making the alphabet a benchmark for computational algebraic geometry approaches."],"supporting_citations":[{"why":"Supplies the canonical (epsilon-form) differential equation framework that turns the master integrals into uniform-weight iterated integrals.","marker":"[32]"},{"why":"Provides the top-sector uniform-weight basis for the double-pentagon family, including the six-dimensional double-pentagon integral whose reduction coefficients are the one allowed epsilon-pole exception.","marker":"[38]"},{"why":"Computes the dual-conformal double-pentagon integrals Omega_even and Omega_odd, to which the top-sector DP integrals reduce up to O(epsilon).","marker":"[48]"},{"why":"Supplies the five-point one-mass alphabet letters that seed the six-point alphabet.","marker":"[42]"},{"why":"Earlier work deriving canonical differential equations for the other planar two-loop six-particle families (d, e, f) used here.","marker":"[35]"},{"why":"Introduces the symbol and Chen iterated integral formalism in which the solutions are presented.","marker":"[1]"},{"why":"Provides the high-precision numerical evaluation used to validate the iterated-integral solutions at four Euclidean points.","marker":"[55]"},{"why":"Supplies the one-fold integral representation used for the weight-three and weight-four solutions.","marker":"[4]"}],"fun_headline_variants":["Complete function space for two-loop six-particle amplitudes","All planar two-loop six-particle master integrals solved","Two-loop six-particle amplitudes: full analytic solution obtained","Chen iterated integrals fully characterize two-loop six-particle space","Bottleneck removed: two-loop six-particle integrals now analytic"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that every integral in a four-dimensional Yang-Mills amplitude reduces onto the uniform-weight master integrals with coefficients that are finite as $\\epsilon$ goes to zero, except for the six-dimensional double-pentagon integral, whose coefficients may be of order $1/\\epsilon$ but whose integral is of order $\\epsilon^2$; this is verified for a complete set of integral numerators up to six powers of loop momentum on the maximal cut, and for a selection without the cut, but not proven for every possible numerator.","fun_headline_variants_meta":{"raw":{"variants":["Complete function space for two-loop six-particle amplitudes","All planar two-loop six-particle master integrals solved","Two-loop six-particle amplitudes: full analytic solution obtained","Chen iterated integrals fully characterize two-loop six-particle space","Bottleneck removed: two-loop six-particle integrals now analytic"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000991,"raw_usage":{"total_tokens":4208,"prompt_tokens":959,"completion_tokens":3249,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":575,"completion_tokens_details":{"reasoning_tokens":3172}},"tokens_in":575,"tokens_out":3249,"duration_ms":24441,"temperature":1.0,"reasoning_tokens":3172,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:19:16.086399+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A single legitimate Yang-Mills two-loop six-particle numerator whose integration-by-parts reduction onto the uniform-weight basis produces a $1/\\epsilon$ coefficient multiplying an integral that starts at order $\\epsilon^0$ would break the finite-part argument; likewise, a complete amplitude evaluation that requires any alphabet letter or symbol outside the 167 letters and 945 symbols at weight four would falsify the claimed completeness.","supporting_citations":[{"cited_title":"Matijašić, Singularity structure of Feynman integrals with applications to six-particle scattering processes, PhD thesis, Munich U., 2024","cited_arxiv_id":null,"evidence_quote":"Provides the top-sector uniform-weight basis for the double-pentagon family, including the six-dimensional double-pentagon integral whose reduction coefficients are the one allowed epsilon-pole exception."},{"cited_title":"Using differential equations to compute two-loop box integrals","cited_arxiv_id":"hep-ph/0005232","evidence_quote":"Computes the dual-conformal double-pentagon integrals Omega_even and Omega_odd, to which the top-sector DP integrals reduce up to O(epsilon)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the symbol and Chen iterated integral formalism in which the solutions are presented."},{"cited_title":"Zoia, Modern Analytic Methods for Computing Scat- tering Amplitudes: With Application to Two-Loop Five- Particle Processes(Springer Nature, 2022)","cited_arxiv_id":null,"evidence_quote":"Supplies the one-fold integral representation used for the weight-three and weight-four solutions."}],"review_version":1}