Answers are grounded in this paper and cite their sources.
The unfixed real-time Henstock–Kurzweil (HK) functional integral for truncated Yang–Mills removes the choice of a gauge surface, but gauge orbits remain objects of the formulation: that is a Level-One bypass of the Gribov ambiguity. The present paper carries out a Level-Two bypass at finite mode truncation: the Wilson loops W(C) are taken as the fundamental variables, so that the configuration space of the theory is a space of loop data and no gauge field, gauge orbit, gauge-fixing condition or Faddeev–Popov determinant occurs anywhere in the formulation. Nine statements are proved, and all of them are machine checked in Lean 4 without sorry, without admit and without new axioms:(A) the loop algebra of Wilson loops of based closed lattice words is a commutative C-algebra, closed under complex conjugation, whose elements are gauge-invariant functionals, with loop variables bounded by one and with the Mandelstam identity tr(AB) + tr(AB⁻¹) = tr A · tr B holding for SU(2);(B) the truncated Yang–Mills action is the evaluation of an explicit multivariate polynomial of total degree at most one in the plaquette Wilson variables, hence a cylindrical function on loop space supported on the loops of length four;(C) every loop-expressible gauge-invariant functional is a cylindrical function of finitely many Wilson loops;(D) the Wilson-loop integrand Φ(W) exp(iS(W)) is HK-integrable on every bounded box of the loop space for every integration calibration, with value the Lebesgue value;(E) the Level-Two integral is a constant function of the gauge-fixing data, and configurations on one gauge orbit — in particular Gribov copies — are the same point of the loop formulation, so the Gribov ambiguity is structurally absent rather than cancelled;(F) for loop observables the Level-Two integral taken with respect to the push-forward of the Level-One measure coincides with the Level-One unfixed HK integral;(G) the one-plaquette SU(2) check is proved in full, including the closed form (2/π)∫₀^π eiβ cos θ sin²θ dθ = 2J₁(β)/β, obtained analytically by term-by-term integration, with J₁ defined by its everywhere convergent power series.Two further results extend the v1 formulation: (C′) the Mandelstam relations for SU(N), derived in manifestly covariant trace form from the vanishing of the antisymmetriser AN+1 on (C^N)⊗(N+1) together with the rank-N Cayley–Hamilton theorem, and written out explicitly for SU(3); and (H) projective compatibility across mode levels: for a cylindrical loop functional of bounded mode support the finite-M Level-Two integrals coincide, so that they form a projective system whose limit exists as a linear functional — not a measure — on the union of the loop algebras.No continuum limit, no mass gap, no confinement and no Makeenko–Migdal equation is claimed; the Giles reconstruction theorem is used only as a cited black box, and no stack- or sheaf-theoretic refinement of it is attempted. The honest report in §13 records exactly where each proof stops. The paper is accompanied by a complete Lean 4 / Mathlib formalisation (nine files, 138 audited declarations, no sorry, no admit, no new axioms) and by a deterministic Python script that regenerates all numerical checks and all six figures.
No takes yet. Share an insight, caveat, or question.
Yuri N. Berdinsky (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: