# Start a new discussion

## Not signed in

Want to take part in these discussions? Sign in if you have an account, or apply for one below

## Site Tag Cloud

Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.

• CommentRowNumber1.
• CommentAuthorUrs
• CommentTimeMay 29th 2010

I moved the proof of the claim that the Segal-Brylinski “differetiable Lie group cohomology” is that computed in the (oo,1)-topos of oo-Lie groupoids from the entry group cohomology to the entry Lie infinity-groupoid and expanded the details of the proof considerably.

See this new section.

Towards the end I could expand still a bit more, but I am not allowed to work anymore today… :-)

• CommentRowNumber2.
• CommentAuthorUrs
• CommentTimeMay 29th 2010

Towards the end I could expand still a bit more,

Hm, in fact I think I have a gap towards the end. So far it just shows an inclusion of Segal-Brylinski cohomology into the (oo,1)-topos cohomology, not an isomorphisms.

Hm, darn, I need to think more about it…

• CommentRowNumber3.
• CommentAuthorUrs
• CommentTimeMay 31st 2010
• (edited May 31st 2010)

I don’t see the equivalence anymore that I claimed originally, it now seems to me that the intrinsic (oo,1)-topos Lie group cohomology is even finer than the Segal-Brylinski differential cohomology (which refines the “naive” Lie group cohomology).

I adjusted the entries accordingly, the details are at oo-Lie groupoid – Lie group cohomology.

But maybe to highlight the sensitive point within the discussion there:

Brylinski’s differentiable Lie group cohomology is obtained (in paraphrase) by choosing funcorially for each $G^{\times_k}$ a cofibrant replacement $C(\{U^k_i\}) \stackrel{\simeq}{\to} G^{\times_n}$ and then setting

$H^n_{diffr}(G,A) := sPSh(\int^{[k]} \Delta[k] \otimes C(\{U^k_i\}) , \mathbf{B}^n A)$

But even though each $C(\{U^k_i\})$ is cofibrant, the above totalization is, while related by a zig-zag of weak equivalences to $\mathbf{B}G$, not cofibrant. What is a cofibrant replacement for $\mathbf{B}G$ is

$\int^{[k]} \mathbf{\Delta}[k] \otimes C(\{U^k_i\})$

where the fat Delta is the Bousfield-Kan resolution $\mathbf{\Delta}[k] = N([k]/\Delta[op]^{op})$ of the point in $[\Delta^{op}, sSet]_{proj}$. So the “true” intrinsic cohomology $H^n_{true}(G,A)$ of $\mathbf{B}G$ is obtained by mapping out of this bigger guy. By pullback along the Bousfield-Kan map we hence have a natural morphism

$H^n_{naive}(G,A) \to H^n_{diffr}(G,A) \to H^n_{true}(G,A) \,.$

First I thought I had shown that the last map is an iso due to the abelianness of $A$, but I don’t see my argument anymore.