A cobordism is called Lagrangian cobordism when L and L' are Lagrangians in a symplectic manifold (M, w) and W is an embedded Lagrangian in [0,1]* R * M with the property that near the boundary it looks like the products over L and L'. A Lagrangian pseudo-isotopy is a Lagrangian cobordism (W; L, L'), diffeomorphic to a trivial cobordism L*[0,1].

In this talk we will see that under some topological constraint an exact Lagrangian cobordism is Lagrangian pseudo-isotopy. We use Floer homology and the s-cobordism theorem. 

suite et fin par Nicolas de l'introduction aux martingales. le cas des martingales L2.

Le but de cet exposé est de donner quelques éléments à propos d'un problème de diffusion inverse à énergie fixée dans les trous noirs de type de Sitter-Reissner-Nordström qui sont des solutions à symétrie sphérique et électriquement chargées des équations de Einstein-Maxwell.

From Gromov's h-principle for diffeomorphism-invariant open partial differential relations on open manifolds, we derive existence theorems on (pseudo-)Riemannian metrics whose curvatures satisfy certain inequalities. Some typical results: Let $M$ be an open manifold of dimension $n\geq2$. For every real-valued continuous function $f$ on $M$, there is a (possibly incomplete) Riemannian metric with Ricci curvature greater than $f$. Let $\lambda1\leq\dots\leq\lambda{n(n-1)/2}$ and $\eps>0$ be real numbers. If $M$ is parallelizable, then $M$ admits a Riemannian metric $g$ such that pointwise the eigenvalues $\sigma1\leq\dots\leq\sigma{n(n-1)/2}$ of the curvature operator of $g$ satisfy $\lambdai<\sigmai<\lambda_i+\eps$. This is joint work with Marc Nardmann.

Joint work with YankI Lekili.

If we think of CP^3 as the space of triples of points on the sphere then the Chiang Lagrangian is the subspace of triples with centre of mass at the origin. We will see that it has non-vanishing Floer cohomology if and only if the coefficient ring has characteristic 5. This calculation is related to Hitchin's work on Poncelet polygons.

In this talk I will discuss some results and speculations about virtual properties of the fundamental group of compact Kaehler manifolds.

La conjecture de Milnor dit que tout groupe de transformations affines agissant de façon proprement discontinue est virtuellement résoluble. Pour le moment, les seuls contre-exemples connus sont ceux donnés par Abels, Margulis et Soifer en 2002 : ce sont des sous-groupes libres du groupe affine produit semi-direct de $SO(2n+2, 2n+1)$ par $mathbb{R}^{4n+3}$. Nous allons montrer comment adapter cette construction au groupe affine produit semi-direct de n'importe quel groupe de Lie réel semisimple non compact par sa propre algèbre de Lie (pour l'action adjointe).