{"id":"e167c46f-fd2a-4027-9419-a723e06fe4b3","arxiv_id":"2508.09705","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A functional Laplace form of the large-N loop equation, solved with a Gaussian path-integral Green function, reproduces Wilson-loop perturbation theory through order (g²N)², including the three-gluon vertex.","lead":"These 1994 notes rewrite the large-N QCD Wilson-loop equation as a functional Laplace equation on loop space and solve it iteratively with a harmonic-oscillator path integral. The payoff: standard perturbative Feynman diagrams, including the three-gluon vertex, are recovered from one unified loop-space formalism.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Three-gluon vertex step (5.20)→(5.21) is asserted, not derived; the ε→0 limit and the 2/3+1/3 split are unverified.","rationale":"The reader’s weakest assumption—the smearing/ε×ε^{-1} machinery of Sect. 4—is indeed the load-bearing point. My reading of Sect. 5.3 confirms that the crucial passage from Eq. (5.20) to Eq. (5.21) is not a calculation but a claim. The text asserts the 2/3+1/3 split and an integration by parts without displaying the integrals or the boundary terms. Since the whole paper is explicitly old unpublished notes, this gap is not mitigated by external verification in the literature; the derivation itself must be checkable. I also note the paper’s own admission that the remaining order-λ² diagrams are not computed, only asserted to follow. This does not change the reader’s CONDITIONAL verdict: the central physics is almost certainly correct, but the proof as written is incomplete. A concrete symbolic or numerical re-derivation of the ε→0 limit would settle whether the three-gluon vertex really emerges with the stated coefficients and whether the result is smearing-independent. Therefore I mark the verdict UNCHANGED: it remains CONDITIONAL, pending verification of the asserted limit.","tokens_in":30539,"tokens_out":11386,"duration_ms":125119,"concrete_test":"Recompute the ε→0 limit of Eq. (5.20) explicitly for a smooth, non-self-intersecting loop, using stepwise integration over the ordered domains and keeping track of all boundary terms in the integration by parts of the ¨G terms. Verify that the sum equals Eq. (5.21). Then repeat with a different smearing function, e.g. G(σ,σ′)=[1−|σ−σ′|/ε]_+ instead of exp(−|σ−σ′|/ε); reparametrization invariance requires the identical vertex. If the coefficient changes, the ε×ε^{-1} limit is not well-defined and the central claim collapses.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The claimed new result—emergence of the three-gluon vertex at order λ²—is not actually demonstrated. The entire step from Eq. (5.20) to Eq. (5.21) is a verbal assertion: the six terms in braces are said to combine, after “doing the uncertainty ε×ε^{-1},” into the announced 2/3+1/3 split, and the two ¨G terms are said to be transformable “by parts” into a form with three ˙x’s. No intermediate formulas are given, no control is shown for the domains |σ1−σ2|∼ε, |σ3−σ4|∼ε and |σ1−σ4|∼ε, and there is no demonstration that the result is independent of the shape of G. If this limit is not well-defined—for instance if the coefficient of the vertex depends on the smearing function or on the way the contour-ordering θ_c is integrated by parts—then the central claim fails. Moreover, the same subsection ends by stating that the remaining order-λ² diagrams (gluon and ghost loop insertions) “should reproduce” the self-energy insertions, without calculation; thus the abstract’s “reproduced through order (g²N)²” is stronger than what is verified.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reformulates the large-N QCD loop equation as a scalar functional Laplace equation on loop space. A polygon discretization of loop space is introduced, with a specific Toeplitz matrix Q_ij chosen so that the discretized Laplacian tends to the functional Laplacian in the continuum limit; the Green function is represented as a Gaussian path integral. The paper then develops a continuum smearing version (parameter ε) and derives rules for Gaussian averages, including averages involving velocities. Section 5 applies these rules iteratively: to order λ it reproduces the gluon-propagator diagram, and to order λ² it claims to reproduce the two-gluon diagram and the three-gluon-vertex diagram. The manuscript is presented as historical notes from the late 1980s/early 1990s, with recent references added.","tokens_in":30924,"tokens_out":4490,"duration_ms":54287,"significance":"If the derivation in Section 5.3 is completed, the paper would provide a self-contained, ansatz-free derivation of the first nontrivial Feynman diagrams of the Wilson-loop average directly from the loop equation, with no fitted parameters and with the discretization and smearing parameters removed in controlled limits. The order-λ calculation is explicit and clean, and the polygon-discretization construction is well motivated. However, the central new claim—the emergence of the three-gluon vertex—is asserted rather than demonstrated, and the paper explicitly defers the remaining order-λ² diagrams. The significance therefore hinges on whether the missing calculation can be supplied.","major_comments":[{"comment":"The transition from Eq. (5.20) to Eq. (5.21) is the central new result and is not derived. The text states that the first four terms 'result, roughly speaking, in 2/3 of the three-gluon vertex' and that the two ẍ terms are transformed 'by parts', but no intermediate formulas are given, no control is shown for the domains |σ1−σ2|∼ε and |σ3−σ4|∼ε, and no check is made that the ε→0 limit is independent of the smearing function G. Since this ε×ε^{-1} limit is exactly what produces the vertex coefficient and its Lorentz structure, the claimed reproduction of the three-gluon vertex is not established. Please provide the complete calculation or explicitly label the step as a conjecture.","section":"§5.3, Eqs. (5.20)–(5.21)"},{"comment":"The text concedes that the remaining order-λ² terms with two ẋ's/two ξ̇'s, one ẋ/three ξ̇'s, and four ξ̇'s 'should reproduce' the gluon and ghost loop insertions, but no calculation is presented. Consequently the abstract's statement that Feynman diagrams are 'reproduced through order (g²N)²' is stronger than what is actually verified. Either perform these computations or weaken the abstract to state that the three-gluon vertex emerges at order λ² while the self-energy insertions are expected and not shown.","section":"§5.3, final paragraph"},{"comment":"Reparametrization invariance of the ε→0 limit is proved in §4.5 only for the single prototype average in Eq. (4.35). The vertex calculation uses the multi-velocity formulas (4.33), (4.39), and (4.40), together with ordered θ_c integrals, and the same shape-independence of the limit must be demonstrated for those objects. The statement in §4.4 that 'general arguments' guarantee the vanishing of unordered terms is plausible, but it does not by itself show that the coefficient of the three-gluon vertex in Eq. (5.21) is independent of the smearing function. A concrete check for the exponential G of Eq. (4.10) would address this concern.","section":"§4.5 and §4.4"}],"minor_comments":[{"comment":"Typo: 'discterization' should be 'discretization'.","section":"Introduction, p. 3"},{"comment":"Typo: 'whith' should be 'with'.","section":"§2.7, Eq. (2.40)"},{"comment":"The notation 'dG' for the integration variable is confusing; suggest using an integration variable g with dg and writing G explicitly as a function of the separation.","section":"Eq. (4.35)"},{"comment":"The expansion in ordered contour integrals assumes convergence and a regulator for the θ_c integrals; a brief comment on the sense in which this expansion is used would improve readability.","section":"Eqs. (5.2)–(5.3)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is essentially a historical note, but it is being submitted with a claimed new result. The missing calculation in §5.3 is the decisive issue: if the author can supply it, the paper would be a useful contribution; if not, the abstract and conclusions should be correspondingly weakened. The old-note character is acceptable if the journal welcomes methodological/historical papers."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a 1994 write-up of 1988-94 work, posted unedited, and its value is genuine but narrower than the abstract claims. The genuinely new item—the emergence of the three-gluon vertex from inverting the loop-space Laplacian—is not actually shown. The step from (5.20) to (5.21) is a summary: six terms 'combine', two G-double-dot terms are transformed 'by parts', and out comes the vertex. No intermediate formulas, no control over the ε-scale domains, no demonstration of smearing-function independence. The stress-test note has it right. That's a load-bearing gap for the central claim.\n\nCredit where it's due: the reformulation of the loop equation as a scalar functional Laplace equation is clean, the polygon discretization with the q-matrix and its Green function is a real piece of work (mostly from ref. [22], but here it's nicely assembled), and the Gaussian averaging rules with velocities in Sect. 4 are useful and clearly stated. The order-λ step is shown in detail—type-2 and type-3 terms vanish, and you see the gluon propagator emerge from the ε × ε^{-1} uncertainty. For that part, the iterative method demonstrably works.\n\nOther soft spots, in proportion: the abstract says 'reproduced through order (g²N)²', but Sect. 5.3 explicitly ends with the gluon and ghost loop insertions 'should reproduce' the self-energy diagrams—that's a promise, not a result. Reparametrization invariance for the full set of vertex terms is not proven; only a single prototype average is checked in Sect. 4.5. And the ε→0 limit is argued by order counting, which is plausible but not rigorous.\n\nThe paper is honest about what it is: unedited notes with a 2005 addition. The citation pattern is fine; the self-citations are to the papers where the pieces actually appeared. It's not a rigorous proof of the three-gluon vertex claim, but it's a serious methodological document with enough substance to justify a careful referee—mainly to try to verify that crux step or at least constrain whether it's true. Who gets value: people working on the loop equation/lattice bootstrap who want the old toolkit, and historians of the loop-space program.\n\nRecommendation: send it to peer review, but the referee should be asked to check (5.20)→(5.21) specifically. If the author can supply the missing calculation, it becomes an important reference; if not, it should be accepted as a technical/historical note with the claim softened.","headline":"Old notes, one new claim, and the crux step is handed over with a wave: worth refereeing for the record, not for the claimed proof.","tokens_in":31332,"tokens_out":2154,"would_cite":false,"duration_ms":22452,"reading_group":"maybe","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 shows that iterating the functional Laplace form of the large-N QCD loop equation reproduces the perturbative Wilson-loop diagrams through order (g^2N)^2, including the three-gluon vertex.","keywords":["loop equation","Wilson loop","large-N QCD","functional Laplacian","loop space","three-gluon vertex","perturbative diagrams","harmonic oscillator path integral"],"falsifier":"Perform the order-$\\lambda^2$ iteration with a generic admissible smearing function $G$ rather than the exponential kernel, keeping all velocity-correlation terms at finite $\\epsilon$. If the resulting $\\epsilon\\to0$ limit depends on the shape of $G$, or if the unordered velocity pairs produce a surviving $\\epsilon\\times\\epsilon^{-1}$ piece, the reconstruction of the three-gluon vertex fails. The direct cross-check is to compare expression (5.21) with the standard Yang-Mills three-gluon amplitude in $d\\neq4$ dimensional regularization.","tokens_in":30399,"feed_emoji":"🌀","tokens_out":7862,"duration_ms":78545,"temperature":0.7,"pith_summary":"The paper is trying to establish that the loop equation of large-N QCD, rewritten as a functional Laplace equation on loop space, can be solved by iteration in the 't Hooft coupling without any ad hoc ansatz for the Wilson-loop average. The solution is organized through Gaussian averages over random loop fluctuations, mediated by a smeared functional Laplacian whose Green's function is a path integral of the Euclidean harmonic oscillator. Iterating once reproduces the gluon-propagator contribution; iterating twice reproduces the two-gluon diagram and the three-gluon vertex diagram. The three-gluon vertex is the paper's new result: it emerges from a nonzero epsilon-times-1-over-epsilon limit of ordered velocity averages, not from inserting a vertex by hand. If the derivation is right, standard planar perturbation theory is an output of the loop equation, and the same functional-Laplace language offers a nonperturbative starting point.","feed_headline":"Three-gluon vertex emerges from iterating the loop equation","feed_subtitle":"Large-N Yang-Mills perturbation theory is reproduced, order by order, from a functional Laplace equation with no hand-inserted vertex.","key_machinery":"The central object is the smeared functional Laplacian $\\Delta^{(G)}$, defined with a Gaussian kernel $G(\\sigma,\\sigma')=e^{-|\\sigma-\\sigma'|/\\epsilon}$ (or its reparametrization-invariant version). At finite $\\epsilon$ its Green's function is a Gaussian path integral whose action is that of an Euclidean harmonic oscillator at finite temperature. The load-bearing identities are the velocity averages (4.33), (4.39), and (4.40): they tell how ordered contour integrals involving $\\dot\\xi(\\sigma)$ are averaged against the exponential measure, and they produce the $1/\\epsilon$ factors. The $\\epsilon\\times\\epsilon^{-1}$ limit of these averages converts the smeared, reparametrization-noninvariant e","core_discovery":"Restated on the paper's own terms, the claim is this: the scalar loop equation\n$$\\$\\Delta$ W(C) = \\$\\lambda$ \\oint_C dx_\\mu \\oint_C\\!\\!\\!\\!\\not\\; dx'_\\mu\\, \\$delta^{{(d)}}$(x-x') W(C_{xx'})W(C_{x'x})$$\nhas an explicit iterative solution\n$$W[x]=1-\\frac12\\int_0^\\infty dA\\left(\\langle J[x+\\sqrt A\\,\\xi]\\rangle_\\$xi^{{(G)}}$-\\langle J[\\sqrt A\\,\\xi]\\rangle_\\$xi^{{(G)}}$\\right),$$\nwhere the average over $\\xi$ is Gaussian with kernel $G(\\sigma,\\sigma')=e^{-|\\sigma-\\sigma'|/\\epsilon}$. Through order $\\lambda$ this reproduces the diagram with a single gluon propagator, and through order $\\lambda^2$ the two-gluon diagram together with the diagram containing the three-gluon vertex. The derivation handles the removal of t","pith_inferences":["The paper demonstrates only orders $\\lambda$ and $\\lambda^2$; a natural extension is to prove that the $\\epsilon\\times\\epsilon^{-1}$ mechanism yields the standard Yang-Mills vertices at every loop order, which would turn this into a derivation of planar perturbation theory from the loop equation.","The Gaussian measure behind the Green's function suggests a stochastic interpretation of Wilson-loop averages: the loop evolves in a proper time $A$ under harmonic-oscillator fluctuations, which could be simulated numerically on discretized loop space.","The scalar-loop example in the paper is a clean test bed: applying the same iterative machinery to the Abelian case should reproduce the exponentiated photon propagator with no three-gluon vertex, isolating the non-Abelian contribution.","The smeared Laplacian may be useful for computing cusp anomalous dimensions, where ordered contour integrals and velocity insertions play the same role as in the three-gluon vertex derivation."],"forward_implications":["The same iteration can in principle be pushed to higher orders in $\\lambda$: the momentum-space expansion in ordered contour integrals and the Gaussian-average formulas are all-order identities, so planar diagrams are generated systematically.","The three-gluon vertex result shows that the smearing method captures genuinely non-Abelian interactions, not just Abelian-like propagator exchanges.","Because the polygon discretization provides a finite-dimensional Laplacian whose continuum limit is a well-defined Gaussian path integral, it supplies a fully loop-space regularization of the Wilson-loop dynamics.","In dimensional regularization the spurious non-reparametrization-invariant terms vanish identically for any $\\epsilon$; with a cutoff they vanish for smooth contours as $\\epsilon\\to0$, so the scheme is compatible with two regularization strategies.","The same functional Laplacian applies to scalar Wilson loops and to supersymmetric loops, so the iterative solution extends beyond pure gluodynamics."],"supporting_citations":[{"why":"Defines the original vector loop equation for large-N QCD whose contour integration this paper converts into the scalar functional Laplace equation.","marker":"[1]"},{"why":"Establishes the equivalence between the vector and scalar loop equations and the identity needed for integrating over the contour.","marker":"[2]"},{"why":"First identifies the loop equation as a functional Laplace equation, the formulation the paper develops.","marker":"[16]"},{"why":"Introduces the polygon discretization of loop space and the Q_ij Laplacian whose continuum limit is the harmonic-oscillator path integral.","marker":"[22]"},{"why":"Provides the stochastic-quantization derivation and regularization of the loop-space equation that motivates the smeared Laplacian.","marker":"[17]"},{"why":"Derives the second-order Schwinger-Dyson equation for the Wilson loop, the field-space counterpart behind Eq. (2.18).","marker":"[20]"},{"why":"Supplies the functional calculus and Gaussian-average framework that the smearing construction extends to velocity-dependent loop functionals.","marker":"[4]"}],"fun_headline_variants":["Iterating the loop equation yields the three-gluon vertex","Loop equation iteration reproduces QCD Feynman diagrams","Three-gluon vertex from iterative solution of loop equation","Functional Laplace equation iterated to get QCD vertices","Loop equation solved without hand-inserting vertices"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The derivation rests on the assumption that, as the smearing parameter epsilon tends to zero, the ordered velocity averages that build the three-gluon vertex are well defined and reparametrization invariant through the cancellation of epsilon-times-1-over-epsilon contributions — a property checked explicitly for one prototype average in Eq. (4.35) but not for the full set of terms entering the vertex.","fun_headline_variants_meta":{"raw":{"variants":["Iterating the loop equation yields the three-gluon vertex","Loop equation iteration reproduces QCD Feynman diagrams","Three-gluon vertex from iterative solution of loop equation","Functional Laplace equation iterated to get QCD vertices","Loop equation solved without hand-inserting vertices"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000212,"raw_usage":{"total_tokens":1210,"prompt_tokens":658,"completion_tokens":552,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":402,"completion_tokens_details":{"reasoning_tokens":483}},"tokens_in":402,"tokens_out":552,"duration_ms":5164,"temperature":1.0,"reasoning_tokens":483,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T20:52:39.735750+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform the order-$\\lambda^2$ iteration with a generic admissible smearing function $G$ rather than the exponential kernel, keeping all velocity-correlation terms at finite $\\epsilon$. If the resulting $\\epsilon\\to0$ limit depends on the shape of $G$, or if the unordered velocity pairs produce a surviving $\\epsilon\\times\\epsilon^{-1}$ piece, the reconstruction of the three-gluon vertex fails. The direct cross-check is to compare expression (5.21) with the standard Yang-Mills three-gluon amplitude in $d\\neq4$ dimensional regularization.","supporting_citations":[{"cited_title":"Makeenko and A.A","cited_arxiv_id":null,"evidence_quote":"Defines the original vector loop equation for large-N QCD whose contour integration this paper converts into the scalar functional Laplace equation."},{"cited_title":"Makeenko and A.A","cited_arxiv_id":null,"evidence_quote":"Establishes the equivalence between the vector and scalar loop equations and the identity needed for integrating over the contour."},{"cited_title":"Gervais and A","cited_arxiv_id":null,"evidence_quote":"First identifies the loop equation as a functional Laplace equation, the formulation the paper develops."},{"cited_title":"Makeenko, Polygon discretization of the loop-space equation , Phys","cited_arxiv_id":null,"evidence_quote":"Introduces the polygon discretization of loop space and the Q_ij Laplacian whose continuum limit is the harmonic-oscillator path integral."},{"cited_title":"Halpern, Yu.M","cited_arxiv_id":null,"evidence_quote":"Provides the stochastic-quantization derivation and regularization of the loop-space equation that motivates the smeared Laplacian."},{"cited_title":"Marchesini, A somment on the stochastic quantization: the loop equation of gauge theory as the equilibrium condition , Nucl","cited_arxiv_id":null,"evidence_quote":"Derives the second-order Schwinger-Dyson equation for the Wilson loop, the field-space counterpart behind Eq. (2.18)."},{"cited_title":"L´ evy,Probl` emes concrets d’analyse fonctionnelle, Paris 1951","cited_arxiv_id":null,"evidence_quote":"Supplies the functional calculus and Gaussian-average framework that the smearing construction extends to velocity-dependent loop functionals."}],"review_version":1}