{"id":"88bf87a8-1b69-448e-b754-274ff8193755","arxiv_id":"2412.17974","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The periodic Wilson loop duality is extended beyond one loop, yielding the two-loop n-particle MHV form factor integrand and new five- and six-particle symbols.","lead":"A single-author preprint computes two-loop maximal-helicity-violating form factors in planar N=4 super Yang-Mills using a periodic light-like polygon Wilson loop. It explains a previously mysterious square root as coming from path ordering across periodic images, and reports new five- and six-particle results.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The two-loop integrand (5.5) depends on a one-period projection rule that is stated by example rather than derived; unless the Lagrangian-insertion construction selects exactly the XC image and the P' multiplicities used here, the n-particle integrand and the two-loop square-root alphabet are not…","rationale":"The paper is a serious computational extension and contains real independent checks: recovery of the known 4-particle remainder, collinear limits for 5- and 6-particle form factors, a 2D limit, and ancillary files with explicit symbols and diagram results. These checks support the 5- and 6-particle results but do not establish the general n-particle integrand because the load-bearing one-period projection is not derived at two loops. The reader identified the compactification and projection as the weakest assumption; I agree and sharpen the concern: the infinite image sum is derived, while the projection to one-period diagrams is the under-specified step. The choice of which images survive controls the XC square roots and the P' subtraction in eq. (5.1), so eq. (5.5) is not fully determined by the periodic Wilson loop observable alone without that derivation. The recommended verdict remains CONDITIONAL: the central claim should be accepted only after the projection rule is derived from the Lagrangian insertion / T-duality construction, or after a direct two-loop Lagrangian-insertion computation confirms the projected XC diagram and the P' combinatorics.","tokens_in":25750,"tokens_out":6219,"duration_ms":65211,"concrete_test":"Compute the two-loop X-type contribution directly from the Lagrangian insertion formula: expand the periodic Wilson loop with two insertions of the chiral Lagrangian L(z) integrated over one period, contract the fields to the topology of eq. (4.21), and compare the resulting integration domain and numerator with the projected XC diagram. If the direct computation does not reproduce exactly z4 < z1+q with coefficient ((p1+p2)^2)^2 and the P' multiplicity used in eq. (5.1), the projection rule is incorrect; if it does, the main residual concern is settled.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that eq. (5.5) is the two-loop n-particle MHV form-factor integrand obtained from the periodic Wilson loop. The derivation has two steps: replace the propagator by the infinite image sum D_xi(x) = sum_n D_xi(x+nq) (eq. 2.5), and then project to diagrams 'that can be contained in one period.' The first step is derived from compactifying the q direction, but the second is not derived at two loops. At one loop the projection is justified via the Lagrangian insertion formula in Section 2.2; at two loops the paper explicitly says it uses the old-fashioned expansion and only 'remembers' that one must project. This projection is exactly what selects the XC diagram, with the ordering z4 < z1+q, and suppresses all other images of the same topology. It also fixes the overcounting subtraction P' in eq. (5.1), which is illustrated only for n=5 ('for example... there are only 12 tuples, not all 14'). Since the XC diagram is the sole source of the new two-loop square roots, and since gauge invariance is demonstrated only for the infinite sum, a different but equally natural projection convention (retaining both +/-q images, or imposing all edge labels to lie in one period before path-ordering) would change the coefficients in eq. (5.5) and alter the square-root letters. The n=4 recovery and the collinear limits are nontrivial consistency checks, but they do not uniquely determine the n-dependent combinatorics of the projection. Thus the n-particle integrand is conditional on the two-loop projection being the one dictated by the compact-direction integration in the Lagrangian insertion formula.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates the weak-coupling expansion of the periodic light-like Wilson loop proposed to be dual to MHV form factors in planar N=4 super Yang-Mills theory. It argues that the periodicity of the contour forces path ordering between points that lie on different edges but in the same periodic image, and it identifies the resulting diagrams with the non-planar square-root contributions that appear in two-loop form factors. The paper gives an n-particle two-loop integrand for the remainder function in eq. (5.5), proves the cancellation of infrared divergences by summing the divergent parts, recovers the known four-particle result, and presents symbols for the two-loop five- and six-particle form factors together with collinear and 2D-limit checks. It also proposes a last-entry condition and a conjecture for the 2D alphabet.","tokens_in":26087,"tokens_out":6941,"duration_ms":70121,"significance":"If the construction is correct, this is a substantial advance: it moves the form-factor/Wilson-loop duality beyond one loop, provides the first two-loop five- and six-particle form-factor results, and gives a structural explanation of the appearance of two-loop square roots. The paper is commendably concrete in several respects: there are no fitted parameters, the integrand is benchmarked against the known four-particle remainder function, the divergence cancellation is shown explicitly in eq. (5.4), and machine-readable ancillary files are supplied for the symbols and diagram results. The main risk is that the one-period projection rule is not derived at two loops, and the central integrations are largely delegated to ancillary files, so the new results cannot be fully verified from the text alone.","major_comments":[{"comment":"The derivation of the n-particle integrand depends on the rule 'take the infinite periodic sum and then project to diagrams that can be contained in one period.' At one loop this rule is justified via the Lagrangian insertion formula, but at two loops the paper explicitly uses the old-fashioned expansion and only 'remembers' the projection. This projection is load-bearing: it selects the XC diagram with the ordering z4 < z1+q, fixes the overcounting subtraction P' in eq. (5.1), and thereby determines the two-loop square-root letters and all relative coefficients in eq. (5.5). The gauge-invariance argument in Section 2.2 applies to the infinite unprojected sum, not to the projected finite integrand. A different but equally natural projection convention would change the coefficients in eq. (5.5). I request a derivation of the two-loop projection from the Lagrangian insertion formula or an equivalent first-principles argument, together with a check that the projected integrand is gauge invariant.","section":"Section 2.2 and Section 5.1, eq. (5.5)"},{"comment":"The central new results, R_5 and R_6, are obtained by integrating the star, curtain, X and XC diagrams, but the paper states only that the integrations are 'straightforward' or 'tedious but straightforward' and places the results in ancillary files. The text does not give the integrated expressions, nor does it spell out how the O(1) regularization-dependent parts cancel beyond the M± divergence sum in eq. (5.4). As a result, the new symbols cannot be verified from the paper alone. I ask the authors to include the integrated results at least for representative configurations, or to provide a reproducible derivation, and to state explicitly which parts of the finite remainder are checked to be independent of the cutoff regularization.","section":"Section 4 and Section 5.2"},{"comment":"The removal of overcounting encoded in P' is defined only through an example for n=5, where 12 tuples survive instead of 14. For general n there is no prescription for which periodic images of the double product IO(1,k)IO(j,l) are identified with each other, and the cyclic summation notation for that term is ambiguous. Because these product terms contribute to the finite symbol, a general and unambiguous definition of the surviving tuples is needed before eq. (5.5) can be used for arbitrary n.","section":"Section 5.1, eq. (5.1)"}],"minor_comments":[{"comment":"The condition on the discrete momentum is written as 'exp(ik·q) = 0 or k·q = 2πn'; the first expression should be 'exp(ik·q) = 1'.","section":"Section 2.2, after eq. (2.5)"},{"comment":"The notation '(···)' for the divergence subtraction and the symbol 'cyclic' are not defined explicitly. Please spell out the range of cyclic images and the precise content of the subtracted term.","section":"Section 5.1, eq. (5.1)"},{"comment":"The statement that a star diagram has eight one-loop triangle square roots, with seven algebraic letters per square root, is not demonstrated in the text. A short derivation or an explicit example would help the reader understand the counting.","section":"Section 4.3"},{"comment":"There are several typos and grammatical slips: 'at weaking coulping' should be 'at weak coupling', and the sentence in the Introduction 'It is first argued from the string picture [18] that should be dual to...' is missing a subject.","section":"Introduction and Section 5.2"},{"comment":"The OPE variables for n>6 are not defined, although the 2D conjecture in eq. (5.8) refers to general n. A short comment on how the parametrization generalizes would be useful.","section":"Appendix B"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the unproven one-period projection rule, which is load-bearing for the square-root alphabet and the coefficients in eq. (5.5). If the authors can derive the two-loop projection from the Lagrangian insertion formula or benchmark the n=5 result against an independent Feynman-integral computation, I would be inclined to accept. The reliance on ancillary files for essentially all integrated results also makes independent verification difficult; I would encourage the editor to request a version with more of the key integrations shown explicitly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is the first real two-loop test of the periodic Wilson loop duality, and it produces new 5- and 6-particle form-factor symbols. Read it if you care about form factors, Wilson loops, or the square-root alphabet.\n\nWhat is actually new: the observation that periodicity forces path ordering between points on different edges, and that this is exactly what generates the non-planar square roots; the n-particle two-loop integrand (5.5); the computation of the two-loop remainder functions for 5 and 6 particles; and the last-entry condition and 2D alphabet conjectures. The paper also passes the checks you want: it recovers the known 4-particle result, and collinear limits match known form-factor/amplitude results. Ancillary files contain the symbols and regulated diagrams, so the work is reproducible in principle.\n\nSoft spots: the paper is upfront that at two loops it uses the old-fashioned expansion and only 'remembers' the projection to one period (Section 2.2). That projection is the load-bearing step, and the stress-test's worry is legitimate: if a different but equally natural projection convention is chosen, the coefficient of the XC diagram, and hence the square-root letters, could change. That said, the 4-particle recovery and collinear limits are nontrivial and constrain the combinatorics; they don't fully pin it down for all n, but this feels like a request for a precise statement rather than a fatal flaw. The bigger practical issue is that many integrations are left in ancillary files with 'straightforward' signposts; a referee will need to spot-check some integrals and may want an appendix outlining the method for one representative diagram. The gauge-invariance argument is clearly presented, though it relies on the infinite sum rather than the projected integrand; the divergence cancellation is shown via cancellation of the M± pieces, which is fine.\n\nBottom line: this paper deserves serious referee time. The method is important, the data is new, the checks are real. I'd recommend accepting with a request to make the projection rule precise for all n and to show more detail on at least one non-trivial integration. For a reading group, the XC square-root story alone is worth the session.","headline":"A serious two-loop test of the periodic Wilson loop duality with new form-factor symbols; the projection rule needs a precise statement but the paper deserves refereeing.","tokens_in":26593,"tokens_out":4292,"would_cite":true,"duration_ms":39497,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T13","81T18","81T60"],"pacs":["11.15.-q","11.15.Bt","11.30.Pb"],"model":"deepseek-v4-flash","headline":"This paper shows that the periodicity of a Wilson loop explains the non-planar square roots in two-loop MHV form factors and supplies the complete two-loop integrand.","keywords":["periodic Wilson loop","MHV form factor","N=4 super Yang-Mills","two-loop","square roots","path ordering","remainder function","symbol"],"falsifier":"Evaluate the integrand (5.5) at $n=7$ and compare the resulting symbol with an independent two-loop Feynman-integral calculation: if the 7-particle remainder function does not reduce to the 6-particle one in the collinear limit, or if any letter disagrees, the periodic-sum-and-project rule is wrong. The conjectured 2D alphabet in eq (5.8) at $n=8$ is a second decisive check.","tokens_in":25540,"feed_emoji":"🌀","tokens_out":11493,"duration_ms":97830,"temperature":0.7,"pith_summary":"This paper claims that the duality between planar MHV form factors in N = 4 super Yang-Mills theory and periodic light-like polygon Wilson loops survives at two loops, provided the periodicity is treated as a physical compact direction rather than a bookkeeping device. The compactification makes the gluon propagator an infinite sum over periodic images, and after summing, only diagrams that fit inside one period are kept; this projection is what restores gauge invariance at one loop. The same projection forces path ordering between points on different edges of the same period, and the paper argues that this ordering is the Wilson-loop origin of the square roots that appear in non-planar Feynman diagrams with four massless and one massive leg. Including the resulting XC diagrams, the paper assembles the n-particle two-loop integrand, proves the cancellation of all divergences, reproduces the known four-particle remainder function, and evaluates the 5-particle and 6-particle remainder functions as new data. A sympathetic reader would care because it turns a formal strong-coupling duality into a working weak-coupling computational scheme and provides the first higher-multiplicity two-loop form factor symbols.","feed_headline":"Periodicity explains square roots in two-loop form factors","feed_subtitle":"New integrand reproduces the four-particle result and yields 5- and 6-particle remainder functions.","key_machinery":"The load-bearing object is the periodic light-like polygon Wilson loop, whose $q$-direction is compact after T-duality, so the coordinate-space propagator becomes the infinite image sum $D_{\\mu\\nu}(x)=\\sum_n D_{\\mu\\nu}(x+nq)$. This sum makes the one-loop gauge variation vanish and, at two loops, generates the path ordering between points on different edges in the same period; the resulting XC diagram is the mechanism that produces the non-planar square roots. Divergences are isolated by two ancillary diagrams $M^\\pm$ built from the one-loop three-mass triangle, and the non-abelian exponentiation theorem selects the $C_F C_A$ color topologies that survive in the logarithm of the Wilson loop.","core_discovery":"The central claim is that the n-particle two-loop MHV form factor is computed by the periodic Wilson loop through a two-step rule: first sum every two-loop gluon diagram over all periodic images, then project out any contribution that cannot be placed inside one period. The projection is not cosmetic; without it the one-loop expansion is gauge dependent, and with it a new class of diagrams appears in which two points on different edges of the same periodic image are path-ordered. These XC diagrams carry exactly the square roots of the four-massless-one-massive non-planar Feynman integrals, so the paper identifies the Wilson-loop origin of those roots. Assembling all contributing topologies (square, curtain, star, Y, X, and XC) and regularizing cusps by a small mass, the author obtains the integrand (5.5), proves that its divergent parts cancel, verifies that at n=4 it reproduces the known two-loop remainder function, and supplies the symbols of the 5-particle and 6-particle remainder functions, which pass collinear limits to the 4-particle form factor and to MHV amplitudes.","pith_inferences":["Inference: if the periodic-sum-and-project rule is general, the same mechanism should generate non-planar square roots in form factors of other half-BPS operators, where the periodic super Wilson loop replaces the bosonic one.","Inference: the antipodal structure the paper spots in the XC diagram's symbol hints that the periodic Wilson loop may be the natural place to prove antipodal duality beyond symbol level, though the paper only observes the structure at two loops.","Inference: a decisive test the paper does not perform is $n=7$: its integrand (5.5) should yield a symbol that reduces to the 6-particle one in collinear limits and matches an independent Feynman-integral computation."],"forward_implications":["The rule 'sum over periodic images, then project to one period' now has a two-loop proof of principle, so the same rule can be tried at three loops, where new XC-type images will appear.","The paper's counting gives exactly $n(n-3)/2$ one-loop-triangle square roots and $n(n-3)/2$ two-loop XC square roots at $n$ points, predicting the number and type of algebraic letters for any multiplicity.","The last-entry condition found for the two-loop form factors matches the MHV-amplitude condition: only $\\langle i-1,i,i+1,j\\rangle$ can appear in the last entry of the parity-even part.","In the 2D limit the 6-particle symbol has 21 letters, and the paper conjectures the 2D alphabet for all even multiplicities, with the alternating-sign letters appearing only in the third slot."],"supporting_citations":[{"why":"supplies the one-loop periodic Wilson loop duality that this paper extends to two loops.","marker":"[23]"},{"why":"provides the two-loop polygon Wilson loop diagram topologies and the divergence-capturing ancillary diagrams used here.","marker":"[15]"},{"why":"introduced the string/strong-coupling picture of the periodic Wilson loop whose T-duality compactification this paper makes quantitative.","marker":"[18]"},{"why":"supplies the Lagrangian insertion formula and gluon expansion used to build the two-loop Wilson loop diagrams.","marker":"[12]"},{"why":"provides the T-duality treatment with momentum-number projection that underlies the periodic sum and one-period projection.","marker":"[40]"},{"why":"gives the known two-loop 4-particle form factor result against which the integrand (5.5) is checked.","marker":"[33]"},{"why":"identifies the non-planar four-massless-one-massive Feynman integral square roots that the XC diagram must reproduce.","marker":"[37]"},{"why":"non-abelian exponentiation theorem that selects the CFCA color topologies surviving in the logarithm of the Wilson loop.","marker":"[43, 44]"}],"fun_headline_variants":["Periodic Wilson loop explains square roots in two-loop form factors","Two-loop MHV form factors from periodic Wilson loop integrand","Periodicity forces path ordering in new Wilson loop diagrams","Square roots in two-loop form factors traced to periodic images","Two-loop form factor integrand proven finite via periodic Wilson loop"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation assumes that the direction connecting one period to the next is genuinely compact, so that every quantity is a sum over repeated copies of the same picture and only one copy is physical; if that compactification is not the correct way to realize the periodic Wilson loop at weak coupling, the projection step, the cross-edge ordering diagrams, and the square-root explanation do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Periodic Wilson loop explains square roots in two-loop form factors","Two-loop MHV form factors from periodic Wilson loop integrand","Periodicity forces path ordering in new Wilson loop diagrams","Square roots in two-loop form factors traced to periodic images","Two-loop form factor integrand proven finite via periodic Wilson loop"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000395,"raw_usage":{"total_tokens":2038,"prompt_tokens":880,"completion_tokens":1158,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":496,"completion_tokens_details":{"reasoning_tokens":1075}},"tokens_in":496,"tokens_out":1158,"duration_ms":8561,"temperature":1.0,"reasoning_tokens":1075,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T05:08:09.216312+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the integrand (5.5) at $n=7$ and compare the resulting symbol with an independent two-loop Feynman-integral calculation: if the 7-particle remainder function does not reduce to the 6-particle one in the collinear limit, or if any letter disagrees, the periodic-sum-and-project rule is wrong. The conjectured 2D alphabet in eq (5.8) at $n=8$ is a second decisive check.","supporting_citations":[{"cited_title":"Two-Loop Polygon Wilson Loops in N=4 SYM","cited_arxiv_id":"0902.2245","evidence_quote":"provides the two-loop polygon Wilson loop diagram topologies and the divergence-capturing ancillary diagrams used here."}],"review_version":1}