Complex geometry seminar

Ce travail concerne le groupe de Cremona du plan sur un corps parfait, c’est à dire le groupe des applications birationnelles du plan projectif qui sont définies sur ce corps. Nous prouvons que ce groupe est engendré par des involutions.
J’expliquerai la décomposition de telles applications en liens de Sarkisov (applications birationnelles simples entre des espaces fibrés simples) et comment cela donne un ensemble de générateurs du groupe de Cremona. Après, je les décomposerai en involutions, parmi lesquelles on peut mentionner les involutions Geiser et Bertini, et des réflexions d’un groupe orthogonal associé à un espace quadratique.
(Travail en collaboration avec Stéphane Lamy.)

On the one hand, the Littelmann’s path model is a combinatorial tool that describes the representation theory of any (symmetrizable) Kac-Moody Lie algebra, available since 1994. The paths in this model are piecewise linear paths in the finite dimensional real vector space spanned by the fundamental weights. But this vector space together with its affine hyperplanes arrangement is also the standard apartment of an object called the masure, introduced by Gaussent-Rousseau in 2008. The masure is playing the role of the Bruhat-Tits building in the Kac-Moody setting. On the other hand, in the finite dimensional setting, Mirkovic and Vilonen developed a geometric model of the aforementionned representations, by introducing subvarieties in the affine Grassmannian associated to a reductive group, first in 2000. Most of the algebraic information can be derived from the associated polytopes, and there is a bijection between paths and polytopes. In 2014, Baumann, Kamnitzer and Tingley defined the Mirkovic-Vilonen polytopes in the Kac-Moody setting using preprojective algebras. Our goal is to take advantage of the combinatoric/geometric nature of the masure to realize Mirkovic-Vilonen polytopes directly from Littelmann’s paths.
This is a joint work with Stéphane Gaussent.

Il est naturel de regarder des transformations birationnelles du plan, càd des isomorphismes des ouverts de Zariski du plan. Il y en a beaucoup qui sont des involutions et on peut se mettre à les classifier à conjugaison près. Sur le corps des nombres complexes une telle involution possède des courbes fixes rationnelles ou bien une unique courbe fixe irrationnelle. Dans ce dernier cas, les classes de conjugaison des involutions sont à bijection avec les classes d’isomorphismes des courbes fixes. Pas surprenant, ce n’est plus le cas sur le corps des nombres réels…
Je vais motiver la classification dans le cas complexe et ensuite je vais raconter ce qui est connu dans le cas réel.

Comme tous les “séminaires communs de géométrie”, nous aurons de 14h à 14h45 une introduction au sujet de niveau Colloquium, puis de 14h45 à 15h15 une pause thé-gateaux-géométrie, puis de 15h15 à 16h la suite de l’exposé de recherche.

We show the following algebraicity result for a complex projective variety X with big representation of π1 into an almost simple algebraic group: There exists a proper subvariety Z ⊂ X such that for any algebraic curve C, any holomorphic map f : C → X with f(C) not contained in Z is induced from an algebraic morphism. As a corollary, we show that such varieties are pseudo-Brody hyperbolic.

We study the T-equivariant cobordism rings for the action of a maximal torus T on smooth varieties over an algebraically closed field of characteristic zero. The rational T-equivariant cobordism rings of a wide range of examples were computed in recent years including the classes of toric varieties, flag varieties and symmetric varieties of minimal rank using mainly the technique of localisation at fixed points. We seek to extend the known results to any smooth projective (horo-)spherical variety with an action of a maximal torus T. Among others, we obtain explicit presentations for the rational equivariant cobordism rings of odd symplectic Grassmannians IG(k,2n+1). Furthermore, using the self-intersection formula, we are able to compute a wide range of classes in the rational T-equivariant cobordism ring.

Irreducible characters form an interesting basis in the space of of class functions (i.e. functions constant on conjugacy classes) on a finite group G, the goal of harmonic analysis and representation theory is to study properties and applications of that basis.

If G is a reductive p-adic group, such as the group of invertible matrices with p-adic entries, then irreducible characters are known to behave in a regular way not only on conjugacy classes (on  which they are constant) but also on the so called stable conjugacy classes, i.e. the set of elements conjugate over the algebraic closure of the base field (for example, two sheets of a hyperboloid in R^3 are two SL(3,R) orbits inside a single stable orbit). This is studied in the theory of endoscopy in harmonic analysis on p-adic group.

I will give an overview of a long term joint project with Kazhdan and Varshavsky aimed at applying algebraic geometry, including l-adic sheaves, to problems in that theory.

%%%%%%%%%%%%%

Comme tous les Séminaires Communs de Géométrie, cet exposé sera en deux parties : une première partie “colloquium” de 14h à 14h45, puis une partie plus avancée de 15h15 à 16h. Une pause thé-gateaux-géométrie vous est proposée entre les deux exposés.