Computational algebraic number theory with geometric applications

Project leader: Professor Farid HosseiniJafari and Professor Colleen Robles
Project manager: Jin Lee
Team members: Aryan Agarwal, William Anderson, George Hall, Michael Klausner, Lauren Schwartz, Austin Setzler, Benjamin Wang

Manifolds are generalizations of Euclidean space. Some examples (like a sphere) are quite familiar, though others are more exotic. There are meaningful ways to attach additional geometric structures called line bundles to the manifold. For example, one could attach a line through the origin to each point of a sphere. Unlike the sphere, the manifolds naturally occurring from Hodge Theory do not necessarily satisfy a property known as compactness, meaning they may wander off the manifold or in an infinite direction. It is possible to compactify these spaces by gluing on ``boundary components at infinity''. However, the line bundle need not admit a well-defined extension to the compactification. Degenerate behavior may exist around these boundary components, or ``cusps'', which twist the line bundles.

This ``twisting'' or monodromy, behavior is described by the character $\chi_\infty$, an eigenvalue of particular matrices describing a special type of linear transformation on a lattice. A recent work of Green, Griffiths, and Robles demonstrated that, for particular line bundles, $\chi_\infty$ is always a root of unity. Our work was to generate additional meaningful examples analogous to their work and determine the extension behavior of the line bundles, finding the possible values $\chi_\infty$ could take. We constructed examples from various extension fields of $\mathbb{Q}$, quaternion algebras, and lattices associated with K3 surfaces.

This project is partially supported by NSF RTG grant (DMS-2231514) and the final paper can be downloaded:

Download Math+2026.Robles (pdf - 465.11 KB)