1

I am trying to understand the cohomology groups $H^1(\Gamma, \mathbb{Z}^r)$, where $\Gamma$ is a finite (or profinite) group with a continuous action on $\mathbb{Z}^r$ (in my setting, $\Gamma$ will be the Galois group of some extension of nonarchimedean fields).

Does anyone have a resource I can start with?

Most of what I have found (e.g. here) has the free abelian group as the domain, and not codomain of cocycles.

Edit (in response to Eric Wofsey's comment): This is a fair point, so I will try to explain my motivation.

It is known that the category of algebraic tori over a given field $F$ is anti-equivalent to the category of finitely-generated free abelian groups over $F$, equipped with a continuous action of $\operatorname{Gal}(F_s / F)$, where $F_s$ is the separable closure of $F$; passing to finite subextensions $F \subset E \subset F_s$ allows us to consider tori defined over $F$ which split over $E$.

Modulo $F$-isomorphism, the $E$-isomorphism classes of these tori can be shown to correspond with finite disjoint unions of sets in bijection with the set $H^1(\operatorname{Gal}(E/F), \mathbb{Z}^r)$ (where $r$ is the dimension of the torus).

Because there are only finitely many such isomorphism classes (at least, in the examples I have computed), this implies that the groups $H^1(\operatorname{Gal}(E/F), \mathbb{Z}^r)$ are themselves finite; I am trying to understand if there is an intrinsic reason for this.

  • 1
    I don't know of any reason to expect that this is much different from just studying group cohomology in full generality. The fact that your module happens to have an underlying abelian group that is free is not especially relevant. – Eric Wofsey Feb 04 '20 at 23:19

0 Answers0