**Invited Abstracts**

**Sanghoon Baek,**** **University of Ottawa

Annihilators of torsion of Chow groups of twisted spin flags

We discuss the relationship between basic polynomial invariants
and classes of fundamental representations. This provides information
on the torsion of the Grothendieck gamma ?ltration and the Chow
groups of twisted spin flags.

This is joint work with Erhard Neher, Kirill Zainoulline, and Changlong
Zhong.

Vladimir Chernousov, University of Alberta

*Groups of type $F_4$ over regular local rings*

In the talk we discuss the Grothendieck-Serre conjecture and Purity
conjecture for torsors of type $F_4$.

Alberto Elduque, Universidad de Zaragoza

*Gradings on the octonions and the Albert algebra*

Given a grading $\Gamma: A=\oplus_{g\in G}A_g$ on a nonassociative
algebra $A$ by an abelian group $G$, we have two subgroups of the
group of automorphisms of $A$: the automorphisms that stabilize
each homogeneous component $A_g$ (as a subspace) and the automorphisms
that permute the components. By the Weyl group of $\Gamma$ we mean
the quotient of the latter subgroup by the former. In the case of
a Cartan decomposition of a semisimple complex Lie algebra, this
is the automorphism group of the root system, i.e., the so-called
extended Weyl group. A grading is called fine if it cannot be refined.

The fine gradings on the octonions and the Albert algebra over
an algebraically closed field (of characteristic different from
2 in the case of the Albert algebra) will be described, as well
as the corresponding Weyl groups.

**John Faulkner**, University of Virginia

*Weyl images of Kantor Pairs*

$BC_1$-graded Lie algebras correspond to Kantor pairs (including
Jordan pairs) via the Kantor construction. Similarly, $BC_2$-graded
Lie algebras correspond to Kantor pairs with so called short Peirce
(SP) gradings. The Weyl group of $BC_2$ acts on the $BC_2$-gradings
and hence on the Kantor pairs with SP gradings. Sometimes, a Weyl
image of a Jordan pair is not Jordan. For finite dimensional simple
Lie algebras, the Kantor pairs with SP gradings are determined by
certain pairs of subsets of the Dynkin diagram, and it is easy to
compute how these subsets change for a Weyl image. Also, the fundamental
reflection of a Kantor pair with

K-dimension 1 can be constructed by a doubling process on certain
Jordan pairs.

This is joint work with Bruce Allison and Oleg Smirnov.

**Skip
Garibaldi, **Emory
University

*Did a 1-dimensional magnet detect $E_8$?*

You may have heard the rumors that $E_8$ has been detected in a
laboratory experiment involving a 2-centimeter-long magnet. This
is fascinating, both because of $E_8$'s celebrity and because it
is hard to imagine a realistic experiment that could directly observe
a 248-dimensional object such as $E_8$. The purpose of this talk
is to address some of the natural questions, such as: What exactly
did the physicists do? How was $E_8$ involved? Does it make sense
to say that they detected $E_8$? Was it the $E_8$ root system, a
Lie algebra, or a Lie group? And why $E_8$ and not some other type,
like $E_7$ or $E_6$? This talk is based on a joint paper with David
Borthwick.

Nikita Karpenko,
Universite
Pierre et Marie Curie / Institut de Mathematiques de Jussieu

*On $p$-generic splitting varieties*

The talk is based on a joint work with A. S. Merkurjev.

Given a prime integer $p$, we detect a class of $p$-generic splitting
varieties $X$ of a symbol in the Galois

cohomology of a field $F$ such that for any equidimensional variety
$Y$, the change of field homomorphism $\mathrm{CH}(Y)\to\mathrm{CH}(Y_{F(X)})$
of Chow groups with coefficients in integers localized at $p$

is surjective in codimensions $< (\dim X)/(p-1)$. This applies
to projective homogeneous varieties of type $F_4$ and $E_8$.

**Max Knus, **ETH Zurich

*Triality over arbitrary fields and over $\mathbb F_1$*
This is a report on work with V. Chernousov and J.-P. Tignol. Clasical
triality occurs in two different settings, as a geometric property
of $6$-dimensional quadrics and as outer automorphisms of order
$3$ of simple groups of type $D_4$. As already observed by \'E.
Cartan, triality is closely related to octonions. In the first part
of the talk we show that, over arbitrary fields, the classification
of trialities is equivalent to the classification of certain $8$-dimensional
composition algebras. In the second part we discuss in a parallel
way triality over Tits' field with one element.

** **

Mark MacDonald, University of British Columbia

*Essential dimension of the exceptional groups*

In this talk I will survey what is known about the essential dimension
and essential $p$-dimension for each of the exceptional groups,
including some new upper bounds for $F_4$, and simply connected
$E_6$ and $E_7$. I will explain some of the techniques used to find
these bounds, including an analysis of their small linear representations.

Tom De Medts, Ghent University (Belgium)

*Exceptional Moufang quadrangles and $J$-ternary algebras*

Moufang polygons have been introduced by Jacques Tits in order
to describe the linear algebraic groups of relative rank two,

and in fact, one of the main motivations is precisely to get a better
understanding of the exceptional groups.

In particular, there are certain rank two forms of groups of type
$E_6$, $E_7$ and $E_8$, for which the corresponding Moufang quadrangles
have been described by Richard Weiss in terms of so-called quadrangular
algebras.

The explicit construction, however, is rather wild, and requires
a very careful coordinatization of certain vector spaces of dimension

$8$, $16$ and $32$, respectively.

We have found a conceptual way of constructing these quadrangular
algebras, starting from $J$-ternary algebras,

where $J$ is a Jordan algebra of capacity two. In the case of the
Moufang quadrangles of type $E_8$, for instance, this construction
involves the tensor product of two octonion division algebras.

We believe that, in fact, every Moufang quadrangle defined over
a field of characteristic different from $2$ can

be obtained in a similar fashion.

This is joint work with Lien Boelaert.

Raman Parimala, Emory University

R-triviality of certain simply connected groups of type $E_8$

We discuss the $R$-triviality of certain simply connected groups
of type $E_8,2^{6,6}$. This leads to the affirmative solution to
Kneser-Tits problem for this case.

This is a joint work with J.-P. Tignol and R. Weiss.

**Holger P. Petersson**, Fakult\"at f\"ur Mathematik
und Informatik \\FernUniversit\"at in Hagen

*Albert algebras*
Albert algebras belong to a wider class of algebraic structures
called \emph{Jordan algebras}. Originally designed in the early

nineteen-thirties as a tool to understand the foundations of quantum
mechanics, Jordan algebras in the intervening decades have grown

into a full-fledged mathematical theory, with profound applications
to various branches of algebra, analysis, and geometry. Albert

algebras, along with their natural allies called \emph{cubic Jordan
algebras}, form an important subclass whose significance comes to

the fore through the connection with exceptional algebraic groups
and Lie algebras. In order to exploit this connection to the

fullest, a thorough understanding of cubic Jordan algebras in general,
and of Albert algebras in particular, is indispensable. The

primary purpose of my lecture will be to lay the foundations for
such an understanding. More specifically, it will be shown that
the

main concepts of the theory can be investigated over arbitrary commutative
rings. Moreover, a novel approach to the two Tits

constructions of cubic Jordan algebras will be presented that works
in this generality and yields new insights even when the base ring

is a field. We then proceed to describe the basic properties of
Albert \emph{division} algebras, with special emphasis on their

(cohomological) invariants. The lectures conclude with stating and
discussing a number of open problems.

**Arturo Pianzola, **Alberta/Mathematical Sciences

*Serre's Conjecture II, Dessins d'Enfants and Lie algebras of
type $D_4$*

We will describe how Serre's Conjecture II and Grothendieck's Dessins
d'Enfants can be used to classify Lie algebras of type $D_4$ over
complex Laurent polynomial rings in two variables and the complex
projective line minus three points respectively.

**Anne Quéguiner-Mathieu**, Université Paris 13

*Applications of triality to orthogonal involutions in degree
$8$*

Among Dynkin diagrams, the diagram of type $D_4$ is specific in
that it does admit automorphisms of order $3$.

The corresponding simply connected algebraic group is the cover
${\mathrm {Spin}}_8$ of the special orthogonal group ${\mathrm {SO}}_8$.
Inner twisted forms of this group can be viewed as the the ${\mathrm
{Spin}}$ group of some algebraic structure, namely an $8$-dimensional
quadratic form, of even, more generally, a degree $8$ algebra with
orthogonal involution. Because of triality, those degree $8$ algebras
with involution actually come by triple. Hence triality sheds a
particular light on the study of involutions in degree $8$.

The talk will describe several concrete applications of this fact.
For instance, we provide explicit examples of non isomorphic involutions
that become isomorphic after scalar extension to a generic splitting
field of the underlying algebra.

**Richard M. Weiss, **Tufts University

*Moufang Polygons*

A generalized polygon is a bipartite graph whose girth equals twice
its diameter. A generalized polygon is the same thing as a spherical

building of rank~2. Generalized polygons are too numerous to classify,
but Tits observed that the generalized polygons that occur as residues

of thick irreducible spherical buildings of rank at least~3 as well
as the generalized polygons that are the spherical buildings associated
to

absolutely simple algebraic groups of relative rank~2 all satisfy
a symmetry property he called the {\it Moufang condition}. A {\it
Moufang polygon} is a generalized polygon satisfying the Moufang
condition. Moufang polygons were subsequently classified by Tits
and myself.

The Moufang condition is expressed in terms of certain distinguished
subgroups of the automorphism group called {\it root groups}. A
Moufang polygon is uniquely determined by a small set of these root
groups and the commutator relations between them. The classification
of Moufang says that the root groups and these commutator relations
are, in turn, uniquely determined by certain algebraic data. Moufang
triangles (i.e. Moufang polygons of diameter~3), for example, are
classified by alternative division rings and Moufang hexagons by
quadratic Jordan division algebras of degree~3. The exceptional
Moufang polygons---those that come from rank~2 forms of the exceptional
groups---and the algebraic structures classifying them are of particular
interest; these include the Moufang triangles determined by an octonion
division algebra, all the Moufang hexagons and several families
of Moufang quadrangles.

In my first lecture I plan to introduce Moufang polygons and give
some idea of the main steps in their classification. In my second
lecture, I will

focus on on the algebraic structures that arise in the context of
the exceptional Moufang polygons. In the remaining lectures I will
introduce buildings of arbitrary rank and attempt to indicate the
central role that Moufang polygons play in Tits' classification
results for spherical and affine buildings.

**Contributed Talks**

**Hernando Bermudez, **Emory University

*A Unified Solution to Some Linear Preserver*

We obtain a general theorem that allows the determination of the
group of linear transformations on a vector space V that preserve
a polyno-

mial function p on V for several interesting pairs (V; p). The proof
is based on methods from the theory of semisimple linear algebraic
groups, in particular a theorem of Demazure on the automorphism
group of some projective varieties. Along the way we make evident
the connection between the transformations that preserve the polynomial
and those that preserve a set of \minimal" elements of V ,
a connection that had previously been observed for numerous special
cases.

This is a joint work with Skip Garibaldi and Victor Larsen**
**

**Caroline
Junkins, **University
of Ottawa

*The J-invariant and Tits algebras for groups of inner type E6*

A connection
between the indices of the Tits algebras of a split linear algebraic
group G and the degree one parameters of its motivic J-invariant
was introduced by Quéguiner-Mathieu, Semenov and Zainoulline
through use of the second Chern class map in the Riemann-Roch theorem
without denominators. We extend their result to higher Chern class
maps and provide applications to groups of inner type E6.

**John Hutchens, **North Carolina State University

k-involutions of Exceptional Linear Algebraic Groups

**pdf available**** here**

**Timothy Pollio, **University of Virginia

*The multinorm principle *

pdf available here

A finite extension $L/K$ of global fields is said to satisfy the
Hasse norm principle if $K^{\times} \cap N_{L/K}(J_L) = N_{L/K}(L^{\times})$,
where $N_{L/K} \colon J_L \to J_K$ denotes the natural extension
of the norm map associated with $L/K$ to the corresponding groups
of ideles. The obstruction for the Hasse norm principle, which is
often nontrivial, was computed by Tate\cite{Cass} in the Galois
case and by Drakokhrust\cite{Drak} in the general case. Similarly,
a pair of finite extensions $L_1 , L_2$ of $K$ is said to satisfy
the multinorm principle if

$$K^{\times} \cap N_{L_1/K}(J_{L_1})N_{L_2/K}(J_{L_2}) = N_{L_1/K}(L_1^{\times})N_{L_2/K}(L_2^{\times}).$$

Some sufficient conditions for the multinorm principle were given
by H\"urlimann\cite{Hurlimann}, Colliot-Th\'el\`ene--Sansuc\cite{CTS},
Platonov--Rapinchuk\cite{PlR}, and Prasad--Rapinchuk\cite{PR}. These
results assert the validity of the multinorm principle if the extensions
are disjoint (or their Galois closures are disjoint) and one of
the extensions satisfies the usual Hasse norm principle. In my joint
work with Rapinchuk\cite{PoR}, we show that the multinorm principle
always holds for a pair of linearly disjoint Galois extensions (even
if both extensions fail to satisfy the Hasse norm principle). I
will outline the proof of this theorem and gives some additional
results and examples. In particular, I will discuss the situation
for extensions that are not Galois or disjoint, and talk about the
generalization of the multinorm principle for three or more extensions.

References

[1] J.W.S. Cassels, A. Frolich (Eds.), Algebraic Number Theory,
Thompson

Book Company Inc., Washington D.C., 1967.

[2] Colliot-Thelene, Sansuc, Private Communication.

[3] Yu. A. Drakokhrust, On the complete obstruction to the Hasse
principle,

Amer. Math. Soc. Transl.(2) 143 (1989), 29-34.

[4] Hurlimann, On algebraic tori of norm type, Comment. math. Helv.
59(1984),

539-549.

[5] V.P. Platonov, A.S. Rapinchuk, Algebraic Groups and Number Theory,
Academic

Press, 1994.

[6] T. Pollio, A.S. Rapinchuk, The Multinorm Principle for linearly
disjoint

Galois extensions, arXiv:1203.0359v1.

[7] G. Prasad, A.S. Rapinchuk, Local-global principles for embedding
of elds

with involution into simple algebras with involution, Comment. math.
Helv.

85(2010), 583-645.

**Igor Rapinchuk**

*On the conjecture of Borel and Tits for abstract homomorphisms
of algebraic groups*

The conjecture of Borel-Tits (1973) states that if $G$ and $G'$
are algebraic groups defined over infinite fields $k$ and $k'$,
respectively, with $G$ semisimple and simply connected, then given
any abstract representation $\rho \colon G(k) \to G'(k')$ with Zariski-dense
image, there exists a commutative finite-dimensional $k'$-algebra
$B$ and a ring homomorphism $f \colon k \to B$ such that $\rho$
can essentially be written as a composition $\sigma \circ F$, where
$F \colon G(k) \to G(B)$ is the homomorphism induced by $f$ and
$\sigma \colon G(B) \to G'(k')$ is a morphism of algebraic groups.
We prove this conjecture in the case that $G$ is either a universal
Chevalley group of rank $\geq 2$ or the group $\mathbf{SL}_{n, D}$,
where $D$ is a finite-dimensional central division algebra over
a field of characteristic 0 and $n \geq 3$, and $k'$ is an algebraically
closed field of characteristic 0. In fact, we show, more generally
that if $R$ is a commutative ring and $G$ is a universal Chevalley-Demazure
group scheme of rank $ \geq 2$, then abstract representations over
algebraically closed field of characteristic 0 of the elementary
subgroup $E(R) \subset G(R)$ have the expected description. We also
give applications to deformations of representations of $E(R).$

**Changlong Zhong** (University of Ottawa)

*Invariant and characteristic map*

In this talk we consider the characteristic map $c:S^*(\Lambda)
\to CH(G/B)$ with $G$ of type B or D. We show that there is a number
b_d for each d, independent on the rank of the group $G$, such that
$b_d\cdot \subset I^W_a,$ where $I^W_a$ is the ideal of $S^*(\Lambda)$
generated by non-constant $W$-invariant elements. The proof uses
computations of symmetric polynomials and the so-called "Ideal
of generalized invariants".

This is joint work with S. Baek and K. Zainoulline.

** **

Top