Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A K3 in phi4

Francis Brown, Oliver Schnetz

TL;DR

The paper shows that Kontsevich's conjecture on polynomial point-counts for graph hypersurfaces is not universally valid, even among physically relevant φ^4 graphs. It develops the c2-invariant and denominator-reduction toolkit to detect non-Tate phenomena and demonstrates explicit counterexamples at 8 and 9 loops, whose finite-field point counts are governed by modular forms arising from singular K3 surfaces. It further identifies a broad, tractable class of graphs with vertex-width ≤ 3 (including wheels and zig-zags) for which Ψ_G remains polynomial in the Grothendieck ring, aligning with mixed Tate motives, and contrasts this with the non-Tate behavior uncovered in higher-loop examples. The work links arithmetic geometry to perturbative quantum field theory, revealing that the periods and residues of primitive graphs can involve modular forms and singular K3 surfaces, with implications for the motivic Galois action on Feynman integrals.

Abstract

Inspired by Feynman integral computations in quantum field theory, Kontsevich conjectured in 1997 that the number of points of graph hypersurfaces over a finite field $\F_q$ is a (quasi-) polynomial in $q$. Stembridge verified this for all graphs with $\leq12$ edges, but in 2003 Belkale and Brosnan showed that the counting functions are of general type for large graphs. In this paper we give a sufficient combinatorial criterion for a graph to have polynomial point-counts, and construct some explicit counter-examples to Kontsevich's conjecture which are in $φ^4$ theory. Their counting functions are given modulo $pq^2$ ($q=p^n$) by a modular form arising from a certain singular K3 surface.

A K3 in phi4

TL;DR

The paper shows that Kontsevich's conjecture on polynomial point-counts for graph hypersurfaces is not universally valid, even among physically relevant φ^4 graphs. It develops the c2-invariant and denominator-reduction toolkit to detect non-Tate phenomena and demonstrates explicit counterexamples at 8 and 9 loops, whose finite-field point counts are governed by modular forms arising from singular K3 surfaces. It further identifies a broad, tractable class of graphs with vertex-width ≤ 3 (including wheels and zig-zags) for which Ψ_G remains polynomial in the Grothendieck ring, aligning with mixed Tate motives, and contrasts this with the non-Tate behavior uncovered in higher-loop examples. The work links arithmetic geometry to perturbative quantum field theory, revealing that the periods and residues of primitive graphs can involve modular forms and singular K3 surfaces, with implications for the motivic Galois action on Feynman integrals.

Abstract

Inspired by Feynman integral computations in quantum field theory, Kontsevich conjectured in 1997 that the number of points of graph hypersurfaces over a finite field is a (quasi-) polynomial in . Stembridge verified this for all graphs with edges, but in 2003 Belkale and Brosnan showed that the counting functions are of general type for large graphs. In this paper we give a sufficient combinatorial criterion for a graph to have polynomial point-counts, and construct some explicit counter-examples to Kontsevich's conjecture which are in theory. Their counting functions are given modulo () by a modular form arising from a certain singular K3 surface.

Paper Structure

This paper contains 35 sections, 42 theorems, 122 equations, 13 figures.

Table of Contents

  1. Introduction
  2. Points on graph hypersurfaces
  3. Feynman integrals and motives
  4. Results and contents of the paper
  5. Graph polynomials
  6. Matrix representation
  7. General identities
  8. Graph-specific identities
  9. Local structure
  10. The five-invariant
  11. Denominator reduction
  12. The class of $X_G$ in the Grothendieck Ring of Varieties
  13. Linear reductions
  14. Three-valent vertices
  15. Counting points over finite fields
  16. ...and 20 more sections

Key Result

Theorem 1

(Belkale-Brosnan). For every scheme $Y$ of finite type over $\mathrm{Spec}\,\, \mathbb Z$, there exist finitely many polynomials $p_i \in \mathbb Z[q]$ and graphs $G_i$ such that where $[Y]_q$ denotes the point-counting function on $Y$, and $s$ is a product of terms of the form $q^n-q$, where $n>1$. In particular, $[X_G]_q$ is not always polynomial.

Figures (13)

  • Figure 1: The wheels with spokes (left), and zig-zags (right).
  • Figure 2: A three-valent vertex
  • Figure 3:
  • Figure 4:
  • Figure 5:
  • ...and 8 more figures

Theorems & Definitions (90)

  • Theorem 1
  • Proposition 2
  • Theorem 3
  • Conjecture 4
  • Conjecture 5
  • Theorem 6
  • Theorem 7
  • Definition 8
  • Definition 9
  • Proposition 10
  • ...and 80 more