Upcoming talks
The seminar will meet Tuesdays at 4:00pm in 2-151 unless otherwise noted.
Past seminars
05.19.09:
Jennifer French (
MIT).
A comparison of the unstable Adams spectral sequence with the Goerss-Hopkins spectral sequence.
** At 3 PM in room 2-147 ** In the special case of nilpotent, p-complete spaces of finite
type, we show that the unit of the adjunction from a theorem of Mandell induces an isomorphism of spectral sequences. After briefly describing the E2 terms of these spectral sequences, we will analyze the Goerss-Hopkins spectral sequence and ultimately show that the cosimplicial resolution used in the unstable Adams spectral sequence is sufficient for computing the derived functors of derivations of commutative simplicial HFp algebras.
In this talk we will show that the space of almost commuting elements in a compact Lie group splits after one suspension.
05.05.09:
Ricardo Andrade (
MIT).
Effective substitutes for spectral sequences (Hope for the lazy algebraic topologist?).
Usually, spectral sequences are hardly a matter of automatic computation.
Except in simple cases, some extra knowledge is necessary for computing
the differentials... And that knowledge is usually missing!
So (recalling the Serre spectral sequence) what would it take to get a
computer to calculate the homology of the total space of a fibration
knowing at least something about the base and the fibre? I will try to
explain one way to answer this vague question and discuss how useful (or
not) the resulting mechanical procedure actually is.
04.28.09:
Inna Zakharevich (
MIT).
Explorations of scissors congruence.
** At 3:30 in room 4-253 ** While we currently still think of many geometric questions in terms similar to those the Greeks thought in, some of our terminology is drastically different. One of the most basic of these is area. Euclid would never have dreamed of assigning a number to a polygon, and questions of how to compute these numbers would never have entered his head. The question of area was rephrased as a question of length. Given any polygon we can dissect it into subpolygons which can be rearranged into a rectangle one side of which is of unit length. When comparing the areas of two polygons one would simply compare the side lengths of these rectangles.
In modern terms this formulation leads to many questions: can we define volume in such a way? If not, what are the obstructions? What about in higher dimensions? What if instead we looked at oriented polygons? In my talk we will explore the classical solutions to these problems, some modern reformulations, and the new directions that these reformulations lead us.
The first basic example here is the configuration space of unordered k-tuples of distinct points in a space M. When specialized to the case of M given by the complex numbers, these spaces can be identified as the space of classical complex, monic polynomials of degree k which have exactly k distinct roots.
Elementary features of these spaces as well as their connections to spaces of knots, links, and homotopy groups of spheres will be addressed. These topics are based on joint work with R. Budney as well as J. Berrick, Y. Wong and J. Wu.
04.07.09:
Nick Rozenblyum (
MIT).
Stable splittings.
** At 4:30 in room 2-151 **
03.31.09:
Steven Sivek (
MIT).
A spectral sequence for Khovanov homology.
** At 4:30 in room 2-151 ** Khovanov homology is a bigraded invariant of knots and links whose graded Euler characteristic is the Jones polynomial. I'll explain its
construction and give two differentials on the associated chain complex.
This gives us a spectral sequence with several interesting applications:
I'll use it to show that Khovanov homology doesn't say anything new in the
case of alternating knots and construct a geometric invariant related to
the slice genus.
03.17.09:
Anatoly Preygel (
MIT).
A∞ structures in topology: Morse homotopy theory.
The singular chains on a topological space form an E∞
algebra, inducing plenty of familiar structures on cohomology (e.g.,
cup product, Massey products, Steenrod operations). For a compact
manifold M, the cohomology of M may be computed using Morse theory and
we can ask: how much of this structure can we see while staying within
the realm of Morse theory. I'll also try to talk about how this story
(formally) looks like string topology and Gromov-Witten theory.
03.10.09:
Dustin Clausen (
MIT).
The Friedlander-Quillen proof of the complex Adams conjecture.
Let X be a finite complex. Every complex vector bundle over X has an
associated spherical fibration, and this gives a map J: K(X) →
Sph(X) from the complex K-theory of X to its group of stable spherical
fibrations. The Adams conjecture concerns the behavior of J with respect
to the Adams operations in K-theory; more precisely, it states that for
any integer k and class v in K(X), some power of k annihilates J(Ψkv-v).
This conjecture, useful for studying the image of J and hence the stable
homotopy groups of spheres, was first proved by Friedlander following a
sketch of Quillen. One can speciously summarize the argument as follows:
reduce to the case where X is a variety over a finite field, and then use
that Ψp identifies with pullback along the Frobenius map X → X.
The talk will be devoted to explaining this proof, centering on etale
homotopy theory, an algebraic-geometric way of capturing much of the
homotopy type of complex algebraic varieties. Basically no algebraic
geometry will be assumed, but the price may be a lot of "moral" arguments
based on topological intuition. I'll try to keep the statements precise,
though.
For a given compact smooth manifold M we consider the space Emb(M,Rk) of smooth embeddings of M into some large Euclidean space Rk, or rather some geometric variant of it, which is a homotopy invariant of M.
I will explain how Goodwillie's cutting method enables us to understand the homotopy type of this space of emeddings. I will then prove that the rational homology of that space is actually an invariant of the rational homotopy type of M. The proof is based on Kontsevich's theorem on the formality of the little cube operad and Arone's description of the layers of Weiss' orthogonal tower for the space of embeddings. This is a joint work with Greg Arone and Ismar Volic.
02.24.09:
Steven Sam (
MIT).
Topological obstructions in graph theory.
In 1978, Lovász proved the Kneser conjecture (a problem in
combinatorial set theory) by rephrasing the problem in terms of graph
colorings and showing that the obstruction to the existence of a
morphism from a graph to the complete graph lies in the homotopy groups
of an associated simplicial complex. Stronger connections were made by
Babson and Kozlov in 2004 using connectivity and characteristic classes
of some different simplicial complexes. I will discuss some of the
aspects that goes into proving their theorems.
02.17.09:
Martin Frankland (
MIT).
A free resolution of spaces?.
In homological algebra, projective and injective resolutions allow to compute derived functors and do various things. In the non-abelian case, following Quillen et al, we can use homotopical algebra, and more precisely simplicial resolutions. They work for algebraic objects such as commutative rings, but also in a homotopy-theoretic setting like spaces.
In this talk, I will present the simplicial resolution of a space by wedges of spheres. We will look at Stover's original construction, and then at the Dwyer-Kan-Stover E2 model structure on pointed simplicial spaces, where a resolution by spheres is a cofibrant replacement. Some applications will be discussed.
Bousfield later expanded the theory, but we won't get to that.
02.10.09:
Olga Stroilova (
MIT).
The moduli stack of formal groups.
Formal group laws arise in stable homotopy theory in the study of
complex oriented cohomology theories.
Lazard's theorem implies that the fibered category that classifies
formal group laws and isomorphisms between them is a prestack over
affine schemes in the fpqc topology. The associated stack is
known as the moduli stack of formal groups, Mfg. There
is an equivalence of categories between quasi-coherent sheaves over
Mfg and comodules over an appropriate Hopf algebroid.
Sheaf cohomology over this stack can be identified with the E2-term
of the Adams-Novikov spectral sequence.
Such observations suggest that the perspective of stacks is of
interest in this area. The purpose of this talk will be to introduce
basic concepts of stacks, formal Lie varieties over schemes, and
formal groups. There will be some discussion about the properties and
geometry of Mfg, and allusions to a
stack/sheaf-theoretic proof of LEFT.
Sources include works by Goerss, Hopkins, and Vistoli.
12.09.08:
Special topology seminar: Kari Ragnarsson (
DePaul University).
Fusion in the Burnside ring.
In this talk I will present recent work, joint with Radu Stancu, in which we obtain a bijection between saturated fusion systems on a finite p-group S and idempotents in the double Burnside ring of S satisfying a "Frobenius reciprocity relation". (These terms will all be defined in the talk.) The theorem and its proof are purely algebraic, so I will focus attention on implications in algebraic topology, answering long-standing questions on the stable splitting of classifying space and generalizing a variant of the Adams-Wilderson theorem, as well as the obvious implications for p-local finite groups.
12.02.08:
Anatoly Preygel (
MIT).
(A bit on) Cobordism categories.
Cobordism categories of manifolds with various added structure have a right to be fairly complicated: they form the source categories for various topological field theories, they relate to interesting diffeomorphism groups (e.g., the mapping class group), etc. Yet, happily, work of Madsen-Weiss,
Galatius-Madsen-Tillmann-Weiss, and others demonstrate that for a wide class of "various added structure" the homotopy type of these cobordism categories can in fact be explicitly understood.
The goal of this talk is to give an overview of what sorts of cobordism categories such results exist for, of what these results look like, and of what sorts of things go into these results.
11.25.08:
David Jordan (
MIT).
Fusion categories.
Fusion categories are semisimple rigid tensor categories with simple
unit and with finitely many isomorphism classes of simple objects.
They arise quite naturally in the contexts of representation theory
and mathematical physics, yet they may be studied using methods from
homotopy theory.
In this talk, I'll introduce these concepts and
summarize various recent works of Etingof, Gelaki, Nikschych, and Ostrik,
which deal with graded fusion categories, and which allow us to completely
classify fusion categories of Frobenius-Perron dimension pn, pq, pqr, and
pq2. Along the way we'll see a 3-category, a classifying space,
some split orthogonal groups, and some applications of group cohomology.
Disclaimer: So far as we can tell, these notions are essentially disjoint
from those discussed last week by Matt Gelvin, despite a coincidence of
buzzwords.
11.18.08:
Matthew Gelvin (
MIT).
BLO1: A weak Martino-Priddy conjecture.
Martino and Priddy conjectured that two finite groups have homotopic
p-completed classifying spaces if and only if they have the same fusion
data. Whatever this means, the conjecture as stated turns out to be
rather hard to prove, and will not be the subject of this talk.
Instead, we will discuss the work of Broto, Levi, and Oliver, replacing
"fusion data" with another algebraic structure, called
the p-centric linking system. The advantage of this approach is that it
is possible to prove, hopefully within an hour, that the resulting
conjecture is true. The disadvantage is that p-centric linking systems
are much harder to work with in their own right than fusion systems, and
it is in this sense that the resulting conjecture is weaker.
Be that as it may, some result is better than none, so that's what we'll
talk about. What's more, the talk will keep in mind the possibility that
someone might be foolish enough to try to give a talk about the stronger
MP conjecture at some future date. If such a talk were to occur, the
current talk might be considered as a sort of introduction.
11.04.08:
James Pascaleff (
MIT).
Geometric examples of A-infinity structures.
I will describe two examples of A-infinity structures on the
cohomology of a manifold, one coming from Hodge-de Rham theory, and the
other from Morse theory. In each case, the construction is geometric and
involves defining a family of operators parameterized by a suitable family
of planar trees. Time permitting, I'll discuss a connection between these
structures and Lagrangian intersection Floer homology.
10.28.08:
Samik Basu (
Harvard).
Hirzebruch problem 8.
We ask the question: Which homology classes of manifolds over
integers are representable by orientation classes of submanifolds? In an
effort to answer the question we first see that all the homology classes
upto dimension 6 are representable by submanifolds. Also for an
m-dimensional manifold M classes in Hm-1(M) and Hm-2(M) are always
representable. So to get the first possible example of a non representable
class we must look in the 7th homology of 10 dimensional manifolds. I will
discuss an example of a 10 dimensional manifold and an element in H7 which
is not representable.
10.21.08:
Inna Zakharevich (
MIT).
Replacing model categories with simplicial ones.
I will briefly cover two papers by Dan Dugger which provide two different constructions for producting simplicial model category approximations to a model category. The first is a localization of the Reedy model structure on simplicial objects over the category. The second looks at simplicial presheaves over a nicely chosen subcategory of your category. ** Seminar at 3:30 in room 2-105 **
10.14.08:
Eric Wofsey (
Harvard).
Finite topological spaces.
You might ask the question of whether S1 is the
smallest connected noncontractible space, for the appropriate notion
of "smallest". For the stupidest possible notion of "smallest",
namely "having the smallest number of points", the answer is
resoundingly no. In fact, the smallest connected noncontractible
space has only four points! I'll show how this works and how to more
generally use elementary algebraic topology to study an arbitary
finite space, culminating with the construction of a canonical
simplicial model for any finite space. If I have extra time, I may
say a little bit about classifying (strong) homotopy types of finite
spaces or other topics.
10.07.08:
Angelica Osorno (
MIT).
2-groupoids and the classification of homotopy 2-types.
It is known that crossed modules classify connected homotopy 2-types.
What about the non-connected case? We show that homotopy 2-types are
classified by 2-groupoids. This is the obvious generalization, since
a crossed module is essentially the same as a 2-group, that is a
2-groupoid with just one object.
09.30.08:
Nick Rozenblyum (
MIT).
The cotangent complex and formal moduli problems.
09.23.08:
Jennifer French (
MIT).
Mapping spaces of $E_{\infty}$ ring spectra.
Rather than talk about spectra that are algebras over an arbitrary $E_{\infty}$ operad directly, we will talk about the Quillen equivalent category of commutative S-algebras. More generally, we are interested in mapping spaces of commutative R-algebras where R is a commutative S-algebra.
There is a generalization of the universal coefficient spectral sequence or the E-Adams spectral sequence resulting in a Goerss-Hopkins spectral sequence computing the homotopy of a such a mapping space. The goal is to identify the E2 term of this spectral sequence as the André-Quillen cohomology of a simplicial algebra over a simplicial operad. In order to do this, we will talk about the category of simplicial algebras over a simplicial operad and how to make sense of derivations and thus the right derived functors of derivations.
09.16.08:
Sam Isaacson (
Harvard).
Minimal Cisinski localizers.
What does homotopy theory look like when all maps are homotopic? This
talk will be a report on an ongoing project to characterize "toy"
model categories: namely, presheaves of sets over a category with
finite nerve.
09.09.08:
Martin Frankland (
MIT).
Finiteness conditions for CW-complexes.
We can consider different finiteness conditions for a CW-complex, e.g.
having finitely many cells in each dimension, being finite-dimensional, or
being finite. They all imply the analogous conditions on the homology of the
space, but are these algebraic conditions also sufficient? In general, no.
In this talk, I will present a paper of C.T.C. Wall from 1965 exploring such
questions. We will look for necessary and sufficient algebraic conditions
corresponding to different kinds of finiteness for a CW-complex. In
particular, there are conditions to know if a complex X is
finitely dominated. When that happens, the obstruction to finiteness turns
out to be a class in the reduced projective class group of the integral
group ring of $\pi_1(X)$.
Please send mailing list requests and questions to Martin Frankland.
Other useful links: