Skip to the Main Content

Note:These pages make extensive use of the latest XHTML and CSS Standards. They ought to look great in any standards-compliant modern browser. Unfortunately, they will probably look horrible in older browsers, like Netscape 4.x and IE 4.x. Moreover, many posts use MathML, which is, currently only supported in Mozilla. My best suggestion (and you will thank me when surfing an ever-increasing number of sites on the web which have been crafted to use the new standards) is to upgrade to the latest version of your browser. If that's not possible, consider moving to the Standards-compliant and open-source Mozilla browser.

February 5, 2006

Seminar on 2-Vector Bundles and Elliptic Cohomology, III

Posted by Urs Schreiber

Transcript of part 2 of our first session.

2) Algebraic K-theory of Bipermutative Categories


Reminder. If RR is a commutative ring (to be thought of as a ring of functions over some space), then the algebraic K-theory of RR is the space

(1)K(R):=ΩB( nBGL n(R)), K(R) := \Omega B \left( \sqcup_n B \mathrm{GL}_n(R) \right) \,,

where Ω\Omega is the operation of forming the based loop space, while BB is the operation of forming the classifying space. Note that nBGL n(R)\sqcup_n B \mathrm{GL}_n(R) is a topological monoid.

The homotopy groups

(2)K i(R):=π i(K(R)) K_i(R) := \pi_i(K(R))

of this space are the (iith) K-theory groups of RR.


Strategy. Try to mimic this with the ring RR replaced by the bipermutative category V\mathbf{V} of finite dimensional vector spaces.


Definition. [BDR] If \mathcal{B} is a bipermutative category, then the algebraic K-theory of the 2-category of finitely generated free \mathcal{B}-modules is

(3)K():=ΩB( n|GL n()|), K(\mathcal{B}) := \Omega B \left( \sqcup_n \left| \mathrm{GL}_n\left(\mathcal{B}\right) \right| \right) \,,

where |GL n()|\left| \mathrm{GL}_n\left(\mathcal{B}\right) \right| is defined as follows.


Remark. |GL n()|\left| \mathrm{GL}_n\left(\mathcal{B}\right) \right| should be nothing but the geometric realization of the nerve of GL n()\mathrm{GL}_n\left(\mathcal{B}\right), when regarded as a 2-category. During the talk, however, we were not fully sure about some details of this identification.


Definition. Let

(4)[p]={0<1<<p} [p] = \{0 \lt 1 \lt \cdots \lt p\}

be isomorphism classes of finite ordered sets. Consider maps

(5)[p](M 0,1,M 1,2,,M p1,p)(GL n()) p, [p] \mapsto (M_{0,1}, M_{1,2}, \cdots, M_{p-1,p}) \in (\mathrm{GL}_n(\mathcal{B}))^p \,,

i.e. ordered sets of pp matrices in GL n()\mathrm{GL}_n(\mathcal{B}).

(These matrices really are 1-morphisms in the 2-category of KV 2-vector spaces. In this sense these ordered tuples are nothing but ordered tuples of pp composable morphisms.)

Now, for each order-preserving map

(6)f:[p][q] f : [p] \to [q]

we get a map

(7)f * : (GL n()) q (GL n()) p (M 0,1,,M q1,q) (M f(0),f(1),M f(1),f(2),,M f(p1),f(p)), \array{ f^* &:& (\mathrm{GL}_n(\mathcal{B}))^q &\to& (\mathrm{GL}_n(\mathcal{B}))^p \\ && (M_{0,1},\dots,M_{q-1,q}) &\mapsto& (M_{f(0),f(1)}, M_{f(1),f(2)}, \dots , M_{f(p-1),f(p)}) } \,,

where on the right hand side M i,jM_{i,j} for ji+1j \neq i+1 denotes the composition of the matrices with indices in the interval [i,j][i,j]. More precisely, we set

(8)M i,i=I n M_{i,i} = I_n

(the n×nn\times n identity matrix in GL n()\mathrm{GL}_n(\mathcal{B})) and

(9)M i,i+k=M i,i+1M i+1,i+2M i+k1,i+k, M_{i,i+k} = M_{i,i+1} \cdot M_{i+1,i+2} \cdots \cdot M_{i+k-1,i+k} \,,

where, recall, “\cdot” is the matrix multiplication functor.

Next, denote by Δ p\Delta^p the standard pp-simplex and by f *:Δ pΔ qf_* : \Delta^p \to \Delta^q the maps on simplices induced by order-preserving maps of ordered sets. Then, finally, we define

(10)|GL n()|= pΔ p×(GL n()) p/ |\mathrm{GL}_n(\mathcal{B})| = \sqcup_p \Delta^p \times (\mathrm{GL}_n(\mathcal{B}))^p/\sim

where the equivalence relation that we divide out by is

(11)(f *S,A)(S,f *A) (f_* S, A) \sim (S, f^* A)

with SΔ pS \in \Delta^p, A(GL n()) qA \in (\mathrm{GL}_n(\mathcal{B}))^q and f:[p][q]f : [p] \to [q] an order preserving map.


My personal remark. As indicated above, I think what is going on is that we regard GL n()\mathrm{GL}_n(\mathcal{B}) as a 2-category with

- objects being natural numbers

- 1-morphisms nnn \to n being n×n\n\times n matrices with entries in vector spaces (such that the determinant of their dimensions is ±1\pm 1)

- 2-morphisms being matrices of invertible linear maps between the entries of the 1-morphsims.

Then we forget about the 2-morphisms and form the geometric realization of the nerve of the remaining 1-category as usual.

Posted at February 5, 2006 11:55 AM UTC

TrackBack URL for this Entry:   https://golem.ph.utexas.edu/cgi-bin/MT-3.0/dxy-tb.fcgi/740

0 Comments & 1 Trackback

Read the post Seminar on 2-Vector Bundles and Elliptic Cohomology, I
Weblog: The String Coffee Table
Excerpt: Review of the 2-vector approach towards elliptic cohomology. Part I.
Tracked: August 7, 2006 1:11 PM