Here start notes taken during the workshop “Coarse Geometry of Groups”, Lille, june 18-22, 2012. Organizers: C. Drutu (Lille + Oxford), R. Tessera (Lyon).

** Talks I missed on monday, june 18th **

I missed the first day, 5 talks. I tried to reconstruct 3 of them from conversations with speakers. This must be full of mistakes, all my own.

**1. Uri Bader **

Bader defines the Weyl group of a group as follows. Consider a Poisson boundary of (in Furstenberg’s sense). Assume that is ergodic on . Then is the centralizer of in the group of measure preserving transformations of .

Motivating example: a lattice in a semi-simple Lie group . Then turns out to be the Weyl group of , a finite group.

Poisson boundaries form a poset, with if there is a -equivariant map of onto . In the motivating example, this poset corresponds to the poset of parabolic subgroups of . All these constructions are natural under group homomorphisms .

Corollary 1 (Bader-Furman)Lattices in products cannot map to lower rank semi-simple groups.

**2. Rufus Willett **

Willett studies the coarse Baum-Connes conjecture. This relates K-theory of a space and K-theory of its Roe algebra (algebra of operators whose kernels vanish outside a bounded neighborhood of the diagonal). A notion of metric (T) property turns out to be an obstruction for surjectivity of the assembly map. To overcome it, one must modify the norm on Roe’s algebra, and introduce the maximum norm with respect to all unitary representations of the Roe algebra.

Metric (T) property states that the spectrum of a coarse Laplacian (e.g. simplicial Laplacian of Rips’ complex) is bounded below for every unitary representation of the Roe algebra.

Example 1Let be a residually finite group, let be the disjoint union of where has finite index in and the intersection of all is trivial. Then

Proposition 2K-theory does change after change to the max norm if X admits a coarse embedding into a Hilbert space.

Guoliang Yu introduced Property A for metric spaces as a tool to construct coarse embeddings into Hilbert spaces. Conversely, it is hard to negate. Here is a consequence of Property A which is easier to negate.

Definition 3Let be a bounded geometry metric space. is said to be uniformly locally amenable if for all and , there exists such that for any finite subset of , there exists a subset if such that diam and

This is a coarse invariant.

**3. Piotr Nowak **

Nowak has a variant of Garland’s formula, used for proving fixed point properties of groups acting on metric spaces. His formula is less general than the one used previously, in that it applies only to Banach spaces (in fact, reflexive Banach spaces). However, it involves linear Poincaré inequalities, i.e.

where is the (ordinary, linear) average of . Functional analytic techniques can be used to estimate the constant . For instance, interpolation (between and ) gives an estimate for target space .

Theorem 4For groups for which property (T) can be proved using Garland’s formula, a fixed point theorem holds on for in an explicit interval containing 2.

The argument applies as well to non isometric but uniformly bounded (with norm close enough to 1) affine actions. The theorem implies vanishing of cohomology. This shows that no such theorem can hold for all .

Example 2– Groups acting cocompactly on -buildings. – Random groups in the density model. Using Bourdon-Pajot, one gets a lower bound on their conformal dimensions.

**4. Announcements **

– September 24-29: Inaugural conference of CEMPI, European Center for Maths, Physics, Interactions.

Operator algebras: 24-26.

Graphs and groups: 27-29.

– September 2-9: Dynamics of homogeneous flows, Technion.

– September 25-29: CRM Summer school on , Barcelona.