English

Solution of all quartic matrix models

Mathematical Physics 2025-09-26 v4 High Energy Physics - Theory Algebraic Geometry math.MP

Abstract

We consider the quartic analogue of the Kontsevich model, which is defined by a measure exp(NTr(EΦ2+(λ/4)Φ4))dΦ\exp(-{N}\,\mathrm{Tr}(E\Phi^2+(\lambda/4)\Phi^4)) d\Phi on Hermitian N×N{N}\times{N}-matrices, where EE is any positive matrix and λ\lambda a scalar. It was previously established that the large-NN limit of the second moment (the planar two-point function) satisfies a non-linear integral equation. By employing tools from complex analysis, in particular the Lagrange-B\"urmann inversion formula, we identify the exact solution of this non-linear problem, both for finite NN and for a large-N{N} limit to unbounded operators EE of spectral dimension 4\leq 4. For finite NN, the two-point function is a rational function evaluated at the preimages of another rational function RR constructed from the spectrum of EE. Subsequent work has constructed from this formula a family ωg,n\omega_{g,n} of meromorphic differentials which obey blobbed topological recursion. For unbounded operators EE, the renormalised two-point function is given by an integral formula involving a regularisation of RR. This allowed a proof, in subsequent work, that the λΦ44\lambda\Phi^4_4-model on noncommutative Moyal space does not have a triviality problem.

Keywords

Cite

@article{arxiv.1906.04600,
  title  = {Solution of all quartic matrix models},
  author = {Harald Grosse and Alexander Hock and Raimar Wulkenhaar},
  journal= {arXiv preprint arXiv:1906.04600},
  year   = {2025}
}

Comments

31 pages, LaTeX. v2: A problem in v1 specific to dimension 4 is now solved. v3: representation of 2-point function as rational function is added. v4: completely rewritten

R2 v1 2026-06-23T09:50:17.487Z