Accueil |
Programme scientifique |
Informations pratiques |
Formulaire d'inscription |
Formulaire de proposition d'exposé |
Participants inscrits |
Mini cours : « K-théorie tordue et applications » par Jean-Louis Tu (Metz)
Conférences 1H :
Natalia Castellana (Barcelone) : « The permutation representation of a p-local finite group »
David Chataur (Lille) : « Multiplicative structures of Cheeger-Simons theory »
Soren Galatius (Stanford) : « Spaces of manifold »
Jesper Grodal (Chicago & Copenhagen) : « Local-to-global principles for finite groups and classifying spaces »
Pierre Guillot (Strasbourg) : « Invariants cohomologiques et cycles algébriques »
Takuji Kashiwabara (Grenoble) : « $K_*(\Omega^{\infty}\Sigma ^{\infty}X;Z/p)$ as a functor of $K_*(X)$ »
Pascal Lambrechts (Louvain-La-Neuve) : « On the rational homology of spaces of smooth embeddings »
Nathalie Wahl (Chicago & Copenhagen) : « Operads, mapping class groups and TCFT's »
Conférences 45mn :
Christophe Cazanave (Paris Nord & Polytechnique) : « Homotopies entre fractions rationnelles »
Eric Hoffbeck (Lille) : « Un critère de Poincaré-Birkhoff-Witt pour les opérades »
My Ismail Mamouni (Casablanca) : « La conjecture H: Une minoration de la conjecture de la dimension cohomologique »
Wolfgang Pitsch (Barcelone) : « Trivial cocycles and invariants of homology spheres »
Antoine Touzé (Nantes) : « Cohomologie du groupe linéaire à coefficients dans les polynômes de matrices »
Samuel Wüthrich (Lausanne) : « Produits sur les K-théories de Morava »
|
|
|
Mini Cours :
La K-théorie tordue fut découverte par Donovan et Karoubi dans les années 1970 dans le but de mieux comprendre l'isomorphisme de Thom en $K$-théorie. Dans [DK], les auteurs définissent pour tout fibré d'Azumaya A sur un espace X (c'est-à-dire un fibré en algèbres dont la fibre en tout point est M_n(C)) un groupe K^A(X) qui ne dépend à isomorphisme près que de la classe de cohomologie [A] dans H^3(X,Z)_{tors}. Plus tard, J. Rosenberg [Ros] a observé que la théorie se généralise aux fibrés en opérateurs compacts, et donc, d'après les travaux de Dixmier-Douady [DD], à H^3(X,Z) tout entier. Cependant, ce n'est qu'après les travaux de Witten [W] en théorie des cordes et de Freed-Hopkins-Teleman que la communauté mathématique a commencé à s'intéresser de plus près à la théorie.
Dans cette série d'exposés, nous commencerons d'abord à exposer les bases de la théorie telles que dans [DK,AS,TXL,K]. Nous expliquerons ensuite plusieurs approches du caractère de Chern en K-théorie tordue [AS2,MS,TX] et nous verrons que le point de vue de la géométrie non commutative est tout à fait naturel. Enfin, nous énoncerons le théorème de Freed-Hopkins-Teleman [FHT] qui établit un isomorphisme entre l'anneau de Verlinde et la K-théorie tordue équivariante d'un groupe compact agissant sur lui m\^eme par conjugaison.
References:
[AS] M. Atiyah, G. Segal, Twisted K-theory, arXiv:math/0407054.
[AS2] M. Atiyah, G. Segal, Twisted $K$-theory and cohomology, arXiv:math/0510674.
[DD] J. Dixmier, A. Douady, Champs continus d'espaces hilbertiens et de $C^*$-algèbres, Bull. Soc. Math. France 91 (1963), 227-284.
[DK] P. Donovan, M. Karoubi, Graded Brauer groups and K-theory with local coefficients, Inst. Hautes Études Sci. Publ. 38 (1970), 5-25.
[FHT] D. Freed, M. Hopkins, C. Teleman, Loop Groups and Twisted K-Theory II, arXiv:math/051123
[K] M. Karoubi, Twisted K-theory, old and new, preprint.
[MS] V. Mathai, D. Stevenson, On a generalized Connes-Hochschild-Kostant-Rosenberg theorem.
[Ros] J. Rosenberg, Continuous-trace algebras from the bundle theoretic point of view, J. Austral. Math. Soc. Ser. A 47 (1989), 368-381.
[TXL] J.-L. Tu, P. Xu, C. Laurent-Gengoux, Twisted K-theory of differentiable stacks, Ann. Sci. Ens 37 (2004), 841-910.
[TX] J.-L. Tu, P. Xu, Chern character for twisted K-theory of orbifolds, Adv. Math. 207 (2006), 455-483.
[W] E. Witten, D-branes and $K$-theory, J. High Energy Phys. 12 (1998), 19-44.
Conférences invitées :
Natalia Castellana : « The permutation representation of a p-local finite group »
This is a joint work with Assaf Libman (University of Aberdeen). The concept of p-local finite group was introduced by C. Broto, R. Levi and B. Oliver in 2003. It provides a suitable structure for studying p-completed classifying spaces of finite groups (in the sense of Bousfield and Kan). Given two p-local finite groups and a fusion preserving morphism between their corresponding Sylow subgroups, we study the question of extending it to a continuous map between their classifying spaces. In the stable category the situation is completely described by K. Ragnarsson. The result obtained requires the construction of the wreath product of a p-local finite group. The results obtained are used to construct permutation representations of p-local finite groups.
References:
1. C. Broto, R. Levi, B. Oliver, The homotopy theory of fusion systems, J. Amer. Math. Soc. 16 (2003), 779-856.
2. K. Rganarsson, Classifying spectra of saturated fusion systems, Algebr. Geom. Topol. 6 (2006), 195-252.
David Chataur : « Multiplicative structure of Cheeger-Simons theory »
Let $M$ be a smooth manifold, the theory of differential characters of Cheeger-Simons $\{{\cal H}(q)^q(M)\}_{q\geq 0}$ is a refinement of Chern-Weyl theory for higher abelian gauge theories. Differential characters come equipped with a graded commutative product $${\cal H}^{q_1}(q_1)(M)\otimes{\cal H}^{q_2}(q_2)(M)\longrightarrow{\cal H}^{q_1+q_2}(q_1+q_2)(M).$$ This theory enjoys many applications in geometry, topology and mathematical physics. For example in E. Witten's work on higher abelian gauge theory, the product above appears as an action functional of a certain 5-dimensional field theory. Differential characters are just a small part of a family of bigraded groups, the differential cohomology groups $\{{\cal H}(*)^*(M)\}, that mix the cochain complex and the de Rham complex. Differential cohomology was recently studied and generalized by M. Hopkins, I. Singer (see also J. Simons, D. Sullivan and P. Turner). The aim of this talk will be to introduce a nice complex ${\cal C}(*)^*(M)$ that computes differential cohomology. This complex has the advantage to be equipped with a natural and rich algebraic structure (it is an $E_\infty$-algebra). This structure induces Steenrod squares in a very nice way, following I. Gomi we will show how E. Witten's work reduces to the computation of such cohomological operations.
References:
1. J. Cheeger, J. Simons, Differential characters and geometric invariants, Lecture Notes in Mathematics 1167, Springer-Verlag (1985), 50-80.
2. M.J. Hopkins, I.M. Singer, Quadratic functions in geometry, topology and M-theory, J. Diff. Geom. 70 (2005), 329-452.
3. J. Simons, D. Sullivan, Axiomatic characterization of ordinary differential cohomology, arxiv:math/0701077v1.
4. P. Turner, A functorial approach to differential characters, Algebr. Geom. Topol. 4 (2004), 81-93.
5. E. Witten, AdS/CFT correspondence and topological field theory, J. High Energy Phys. (1998), Paper 12.
Pierre Guillot : « Invariants cohomologiques et cycles algébriques »
Il est fréquent qu'un groupe algébrique G soit présenté comme le groupe des automorphismes d'une certaine structure algébrique : ainsi le groupe O_n est lié aux formes quadratiques, G_2 aux algèbres de quaternions, etc. Un invariant cohomologique de G est alors une règle qui associe à l'un de ces objets une classe dans la cohomologie Galoisienne du corps de base. Par exemple, Milnor a montré comment associer à une forme quadratique ses classes de Stiefel-Whitney. Ces invariants ont suscité de nombreux travaux récents. Or il existe une suite spectrale, dite de Bloch et Ogus, qui contient sur sa première page à la fois Inv(G) (l'algèbre de tous les invariants) et CH^*(BG), l'anneau des cycles algébriques sur "l'espace classifiant" de G. De plus, cette suite converge vers la cohomologie (étale) du même espace. Une observation détaillée montre alors que Inv(G) est d'autant plus ``gros'' et intéressant que l'application cycle CH^*(BG) -> H^*(BG) n'est pas un isomorphisme. L'étude de cette application, de manière indépendante, fait également l'objet de multiples publications. Je vais essayer de présenter tout ceci de manière élémentaire, tout en exposant les résultats que j'ai pu avoir.
Jesper Grodal : « Local-to-global principles for finite groups and classifying spaces »
In this talk I will show how one can sometimes "uncomplete" the p-completed classifying space of a finite group, to obtain the original (non-completed) classifying space, and hence the original finite group. This "uncompletion" process is closely related to well-known local-to-global questions in group theory, such as the classification of finite simple groups. The approach goes via the theory of p-local finite groups. This talk is a report on joint work with Bob Oliver.
Soren Galatius : « Spaces of manifolds »
I will present a joint work with Madsen, Tilmann and Weiss. The $d$-dimensional cobordism category has as objects all closed $(d-1)$-manifolds, and as morphisms all compact $d$-dimensional cobordisms. To any category $C$ one can associate a classifying space $BC$, built by first taking a point for each object of $C$, then attaching an edge between $x$ and $y$ for each morphism from $x$ to $y$, then attaching a $2$-simplex for each commutative triangle, and so on. In general, the set of connected components of $BC$ is the set of objects modulo the equivalence relation generated by the morphisms. When $C$ is the bordism category, this gives the set of cobordism classes of $(d-1)$-manifolds, as calculated by Thom. In this talk I will explain how to completely determine the homotopy type of $BC$. As a corollary we get a new proof of Madsen-Weiss' generalized mumford conjecture.
Reference : arXiv:math/0605249
Takuji Kashiwabara : « $K_*(\Omega^{\infty}\Sigma^{\infty}X;Z/p)$ as a functor of $K_*(X)$ »
Is the mod $p$ $K$-theory of $QX$ (where $Q=\Omega^{\infty}\Sigma^{\infty}$) a functor of the $K$-theory of $X$? The traditional answer to this question is no. In this talk we discuss a framework in which we can answer positively to this question, even when $p$-torsion occurs in the integral $K$-theory of $X$.
Pascal Lambrechts : « On the rational homology of spaces of smooth embeddings »
This is a joint work with Greg Arone and Ismar Volic. We prove that the rational homology of the space of smooth embeddings "modulo immersions", $\overline{\mathop{\rm Emb}}(M,R^d)$, of a compact manifold $M$ in a large euclidean space is an invariant of the rational homotopy type of $M$. A special case of this is when $M$ is $1$-dimensional where we get that the homology of this embedding space is the homology of an explicit graph complex (also joint work with V. Turchin). The techniques are Goodwillie-Weiss calculus of embeddings, Weiss orthogonal calculus, and a relative version of Kontsevich's formality of the little disks operad.
References : arXiv:math/0607486, arXiv:math/0703649.
Nathalie Wahl : « Operads, mapping class groups and TCFT's »
Conférences contribuées :
Christophe Cazanave : « Homotopies entre fractions rationnelles »
On munit l'ensemble des classes d'homotopie "algébriques" de fractions rationnelles pointées de degré $n$ à coefficients dans $k$ d'une structure naturelle de monoïde et l'on construit un isomorphisme entre ce monoïde et celui des orbites sous l'action de $SL_{n}(k)$ de formes bilinéaires symétriques non dégénérées sur $k^n$, muni de la somme orthogonale. On mettra ce calcul en rapport avec le calcul des classes d'homotopie d'applications de $P^1$ dans lui-même en homotopie motivique.
Eric Hoffbeck : « Un critère de Poincaré-Birkhoff-Witt pour les opérades »
La notion d'opérade est utilisée pour modéliser des catégories
d'algèbres. Pour certaines bonnes opérades, les opérades de Koszul, on
a un complexe explicite qui permet de déterminer effectivement la
(co)homologie de la catégorie d'algèbres associée et donc de faire des
calculs d'algèbre homologique.
Le but de l'exposé sera de donner un critère effectif pour montrer
qu'une opérade est de Koszul. Notre critère généralise celui introduit
par Priddy pour les algèbres. Priddy définit ce qu'il appelle une base de
Poincaré-Birkhoff-Witt : une base monomiale possédant certaines
propriétés par rappport à une relation d'ordre, assurant qu'une
algèbre est de Koszul. Dans le cadre des opérades, les monômes sont
remplacés par des compositions organisées selon des arbres. L'essentiel
du travail sera d'adapter la notion de base de Poincaré-Birkhoff-Witt à
des structures arborées.
Référence : http://arxiv.org/abs/0709.2286
My Ismail Mamouni : « La conjecture H : Une minoration de la dimension cohomologique »
Les CW complexe finis 1-connexes, notés ci-après $X$, se divisent en deux classes:
- Les CW complexes elliptiques tels que $\dim\pi_*(X)\otimes Q<\infty$.
- Les CW complexes hyperboliques tels que $\dim\pi_*(X)\otimes Q=\infty$.
La cohomologie rationnelle $H^*(X;Q)$ d'un espace elliptique $X$ est à dualité de Poincaré
et sa charactéristique d'Euler, $\chi$ vérifie $\chi\geq 0$.
Bien que les espaces elliptiques jouissent de propriétés très spécifiques (nous en verrons d'autres au cours d'exposé),
ce sont ceux que l'on rencontre le plus couramment en géométrie différentielle (Groupes de Lie, espaces homogènes,...).
Dans cet exposé nous nous intéressons au problème suivant, émis par M.R. Hilali en (1990) :
Soit $X$ un CW complexe elliptique. On a : $\dim\pi_*(X)\otimes Q\leq\dim H^*(X;Q)$.
Cette propriété a été démontrée par M.R. Hillali pour les CW complexes elliptiques tels que $\chi>0$ (cas pur).
Nous démontrerons la conjecture H dans quelques cas où $\chi =0$.
Nous suivrons le plan suivant :
- Quelques points de la théorie des modèles minimaux de Sullivan.
- Enoncé de la conjecture en termes de modèles minimaux.
- Le cas hyper-elliptique, sous condition. (Enoncé et démonstration du théorème).
- Le rang torique. (Enoncé et démonstration du théorème).
- Le cas d'une variété symplectique. (Enoncé et démonstration du théorème).
Wolfgang Pitsch : « Trivial cocycles and invariants of homology spheres »
In 1989 Shigeyuki Morita gave a new construction of the Casson invariant. His construction amounted basically to show that the intersection form of a surface defines a trivial 2-cocycle on the Torelli group of a genus $g$ surface. A well-chosen trivialization of this cocycle plus a correcting term was shown to coincide point-wise with the Casson invariant viewed as a function on the Torelli groups. Recall that the Torelli group is the kernel of the action of the mapping class group on the first homology of the underlying surface. In this talk we will give a general procedure to construct invariants of homology spheres out of trivial cocycles on the Torelli groups. There is a unique obstruction for a trivial cocycles to have an invariant as a trivializing function. This obstruction is shown to be a cohomology class living in the first cohomology group of a precise subgroup of the mapping class group with coefficients in the universal torsion-free abelian quotient of the Torelli group. If one applies this procedure to pull-backs of 2-cocycles defined on the universal torsion-free abelian quotient of the Torelli group, then there is up to a multiplicative constant a unique candidate $J_g$ to produce an invariant. An elementary computation shows that the aforementioned obstruction for the cocycle $2 J_g$ vanishes. We prove from its definition that the invariant thus obtained satisfies the surgery formulas that carachterize the Casson invariant. The presence of the factor 2 reflects the fact that the Casson invariant is an integral lift of the Rohlin invariant.
Antoine Touzé : « Cohomologie du groupe linéaire à coefficients dans les polynômes de matrices »
En caractéristique nulle, la théorie des invariants est relativement bien comprise. En caractéristique positive, la tâche est plus difficile car ces invariants s'enrichissent des invariants supérieurs : les foncteurs dérivés des invariants sous l'action d'un groupe algébrique, ou cohomologie rationnelle, ne sont pas toujours nuls. Dans un article récent Franjou et Friedlander entament l'étude des bifoncteurs strictement polynomiaux. Si $B$ est un tel bifoncteur homogène de bidegré $(n,n)$, les groupes d'extensions $Ext^*(\Gamma^n gl,B)$, ou cohomologie du bifoncteur $B$, calculent la cohomologie rationnelle du groupe algébrique $GL_n$ à coefficients dans $B(k^n,k^n)$. Dans l'exposé, on fera un survol de motivations pour les calculs de cohomologie des bifoncteurs, puis on montrera comment déterminer la cohomologie du bifoncteur $B=S^{n(r)}\mathrm{hom}(-_1,-_2)$, un $r$-twist de puissances symétriques de matrices, d'importance fondamentale dans la théorie.
Samuel Wüthrich : « Produits sur les K-théories de Morava »
Les célèbres théorèmes de nilpotence et de périodicité distinguent les K-théories de Morava comme des objets d'un intérêt central pour l'étude de la catégorie d'homotopie stable. Malgré leur rôle prééminent, des problèmes fondamentaux concernants leur nature restent ouverts, tel qu'une bonne compréhension des structures multiplicatives sur les spectres repr sentants $K(n)$. Je vais présenter un résultat de classification pour des structures de produit homotopique, obtenu en collaboration avec Alain Jeanneret. Nous établissons une bijection naturelle entre les produits sur $K(n)$ et les formes bilinéaires sur un certain $K(n)_*$-module de rang $n$. Comme une conséquence assez surprenante, nous trouvons que l'algèbre des coopérations $K(n)_*(K(n))$ est une invariante complète pour les produits aux nombres premiers impairs.