-
p-adic Borel extension for Shimura varieties and period images
CIRM, September 2026.
Abstract.
In 1972, Borel proved that every holomorphic map from a product of punctured
unit disks to a complex Shimura variety extends to a map from a product
of disks to its Baily--Borel compactification. Recently, Oswal--Shankar--Zhu
and Patel proved the corresponding p-adic statement over discretely valued
fields in the abelian-type case using Rapoport--Zink uniformizations. I will
discuss joint work with Oswal, Shankar, and Yao extending this result to all
Shimura varieties for large primes p by instead relying on techniques from p-adic
Hodge theory. I will also describe how the same proof works in the context of
geometric period images, at least under a good reduction hypothesis.
-
o-minimal GAGA, Hodge theory, and compact moduli of algebraic varieties
ICBS, August 2026.
Abstract.
For any complex algebraic variety X, the singular cohomology of X is an
important topological invariant. This invariant can be enhanced to a
Hodge structure, which additionally records the integrals of algebraic
differential forms along topological cycles, and therefore also sees the
algebraic structure. Hodge structures turn out to be extremely powerful
tools in many areas of mathematics, for example in moduli theory: any family
of algebraic varieties yields a family (or "variation") of Hodge structures,
which can often be used to build a moduli space of those varieties. The
downside is that since Hodge structures are by their very nature highly
transcendental, it is more difficult to realize these moduli spaces as algebraic varieties themselves.
Famously, this story plays out very elegantly for the moduli space Ag of
(principally polarized) g-dimensional abelian varieties where this
perspective gives an exact moduli space, and the work of Baily--Borel
provides a beautiful projective compactification of this space which
connects the Hodge theory to the classical theory of automorphic forms.
I will explain joint work with S. Filipazzi, M. Mauri, and J. Tsimerman
which generalizes this picture to any family by constructing Baily--Borel
compactifications for arbitrary variations of Hodge structures using o-minimal
GAGA. This settles a question of Griffiths first raised in his 1970 ICM address.
Our result has especially nice applications to the moduli of Calabi--Yau varieties,
and in particular implies the b-semiampleness conjecture of Prokhorov--Shokurov.
-
Mixed non-abelian Hodge theory and the linear Shafarevich conjecture
ICM, July 2026.
Abstract.
For a complex algebraic variety X, classical Hodge theory imposes extra structure on
the singular cohomology Hk(X,C) of X using the theory of harmonic forms.
Non-abelian Hodge theory similarly enriches the geometry of the algebraic
variety of representations of the fundamental group (which is thought of
as H1(X,GLr(C))) by using harmonic metrics on local systems. The theory in
the "pure" case of smooth projective X was largely worked out in the 1990s
beginning with the seminal work of Simpson and has seen many applications, for
example the proof of Eyssidieux--Katzarkov--Pantev--Ramachandran of the linear
Shafarevich conjecture for such X. In this talk, I will survey recent advances in
the "mixed" case of arbitrary complex algebraic varieties X, and show how these tools
can be used to prove a number of important results about the topology of algebraic
varieties, as well as a general version of the linear Shafarevich conjecture.
This is based on joint work with Y. Brunebarbe and J. Tsimerman.
-
Deformations of fibered Calabi--Yau manifolds
Princeton, December 2025.
Abstract.
A number of years ago, Kollár proved that any small deformation of an elliptically
fibered Calabi--Yau variety X is also elliptically fibered, provided H2(X,OX)=0.
We show the same is true for any fibration of a Calabi--Yau manifold. More generally, without any
assumption on H2(X,OX), any small deformation of a semiample bundle remains semiample
up to numerical equivalence. I will describe the proof using some Hodge theory and the T1 lifting criterion
of Kawamata--Ran, as well as discuss some related questions. This is joint work in progress with Kristin
DeVleming, Stefano Filipazzi, Radu Laza, Jennifer Li, Roberto Svaldi, Chengxi Wang, and Junyan Zhao.
I'm not teaching Autumn 2026.