Géométrie et Théorie
des Modèles
Année 2020  2021
Organisateurs :
Zoé
Chatzidakis, Raf Cluckers.
Pour recevoir le programme par email, écrivez à :
zchatzid_at_dma.ens.fr. Pour les personnes ne
connaissant pas du tout de théorie des modèles, des notes introduisant les
notions de base (formules, ensembles définissables,
théorème de compacité, etc.) sont disponibles
ici. Elles peuvent aussi
consulter les premiers chapitres du livre Model Theory and Algebraic Geometry, E. Bouscaren ed.,
Springer Verlag, Lecture Notes in Mathematics 1696, Berlin 1998.
Les notes de quelquesuns des exposés sont disponibles.
Vendredi 16 octobre, Zoom.
9h  10h20 : Dmitry Novikov (Weizmann
Institute), Complex Cellular
Structures
Real semialgebraic sets admit socalled cellular decomposition, i.e.
representation as a union of convenient semialgebraic images of standard cubes.
The GromovYomdin Lemma (later generalized by Pila and Wilkie) proves that the maps could be chosen of C^rsmooth norm
at most one, and the number of such maps is uniformly bounded for finitedimensional families.
This number was not effectively bounded by Yomdin or Gromov, but it
necessarily grows as r → ∞.
It turns out there is a natural obstruction to a naive holomorphic complexification of this result related to the natural hyperbolic metric of complex holomorphic sets.
We prove a lemma about holomorphic functions in annulii, a quantitative version of the great Picard theorem.
This lemma allowed us to construct an effective holomorphic version of the cellular decomposition results in all dimensions, with explicit polynomial bounds
on complexity for families of complex (sub)analytic and semialgebraic sets.
As the first corollary we get an effective version of YomdinGromov Lemma with polynomial bounds on the complexity, thus proving a longstanding Yomdin conjecture about tail entropy of analytic maps.
Further connection to diophantine applications will be explained in
Gal's talk.
Notes de l'exposé
et vidéo.
10h30  11h50 : Gal Binayamini (Weizmann Institute), Tame geometry and diophantine approximation
Tame geometry is the study of structures where the definable sets admit finite complexity. Around 15 years ago Pila and Wilkie discovered a deep connection between tame geometry and diophantine approximation, in the form of asymptotic estimates on the number of rational points in a tame set (as a function of height). This later led to deep applications in diophantine geometry, functional transcendence and Hodge theory.
I will describe some conjectures and a longterm project around a more
effective form of tame geometry, suited for improving the quality of the
diophantine approximation results and their applications. I will try to
outline some of the pieces that are already available, and how they
should conjecturally fit together. Finally I will survey some
applications of the existing results around the ManinMumford
conjecture, the AndreOort conjecture, Galoisorbit lower bounds in
Shimura varieties, unlikely intersections in group schemes, and some
other directions (time permitting).
Notes de l'exposé
et vidéo
(les deux premières minutes manquent).
Vendredi 13 novembre :
9h  10h20 : Will Johnson (Fudan U), The étaleopen
topology
Fix an abstract field K. For each Kvariety V, we will define an “étaleopen” topology on the set V(K) of rational points of V. This notion uniformly recovers (1) the Zariski topology on V(K) when K is algebraically closed, (2) the analytic topology on V(K) when K is the real numbers, (3) the valuation topology on V(K) when K is almost any henselian field. On pseudofinite fields, the étaleopen topology seems to be new, and has some interesting properties.
The étaleopen topology is mostly of interest when K
is large (also known as ample). On nonlarge fields, the
étaleopen topology is discrete. In fact, this property
characterizes largeness. Using this, one can recover some wellknown
facts about large fields, and classify the modeltheoretically stable
large fields. It may be possible to push these arguments towards a
classification of NIP large fields. Joint work with ChieuMinh Tran,
Erik Walsberg, and Jinhe Ye.
Notes de l'exposé
et vidéo
(les cinq premières minutes manquent).
Vendredi 27 novembre, 9h : suite et fin de
l'exposé. Notes et vidéo de
l'exposé.
10h30  11h50 : Jinhe (Vincent) Ye
(Sorbonne Université), Belles
paires of valued fields and analytification
In their work, Hrushovski and Loeser proposed the space V̂ of
generically stable types concentrating on V to study the homotopy
type of the Berkovich analytification of V. An important feature
of V̂ is that it is canonically identified as a projective
limit of definable sets in ACVF, which grants them tools from
model theory. In this talk, we will give a brief introduction to
this object and present an alternative approach to internalize
various spaces of definable types, motivated by Poizat's work on
belles paires of stable theories. Several results of interest to
model theorists will also be discussed. Particularly, we recover
the space V̂ is strict prodefinable and we propose a
modeltheoretic counterpart Ṽ of Huber's
analytification. Time permitting, we will discuss some comparison
and lifting results between V̂ and Ṽ. This is a joint
project with Pablo Cubides Kovacsics and Martin Hils.
Vidéo.
Programme des séances
passées : 200607,
200708,
200809,
200910,
201011,
201112,
201213,
201314,
201415,
201516,
201617,
201718,
201819,
201920.
Retour
à la page principale.
