Motives of graph hypersurfaces with torus operations (Q2343154): Difference between revisions

Extensive calculations of Feynman amplitudes revealed that in many cases they evaluate to rational linear combinations of multiple zeta values, which makes them the periods of mixed Tate motives. Kontsevich related the calculation to point counting on a hypersurface defined by the singularities of the Feynman integrand. This paper studies this motive structure for graph hypersurfaces admitting algebraic torus operations. A graph hypersurface \(X_\Gamma\) is the zero set in \(\mathbb{P}^{n-1}\) of a graph polynomial defined as \(P_\Gamma:=\sum_\tau\prod_{e\notin\tau}\!X_e\), where \(\tau\) runs over the spanning trees of \(\Gamma\), and \(X_e\) is the variable assigned to edge \(e\). The paper first gives criteria for the existence of a torus operation on general graph hypersurfaces, and then identifies a class of graphs (so-called \(*\)-graphs), for which this operation is particularly transparent. A motivic version of the Bialynicki-Birula decomposition theorem is then applied to show how the cohomological motive \(M(X_\Gamma)\), in the sense of Voevodsky's triangulated category \(DM(k)\), can be computed for some such graphs when one knows the fixed point loci in a resolution of singularities of \(X_\Gamma\). This is illustrated with only one example however, when \(\Gamma\) is the wheel with three spokes.