3
$\begingroup$

Let $G$ be a p-adic group, and use $\mathcal{H}(G)$ to denote the space of locally constant compactly supported functions on $G$. $\mathcal{H}(G)$ becomes an associative algebra under convolution. We know that nondegenerate modules of $\mathcal{H}(G)$ corresponds to smooth representations of $G$.

Now let $S$ be a right $\mathcal{H}(G)$-module, and $V$ be a left $\mathcal{H}(G)$-module, both are nondegenerate. Consider the tensor product $S\otimes V$. Let $N_1$ to be the subspace of $S\otimes V$ spanned by the elements of the form $$s.f\otimes v- s\otimes f.v, \forall f\in \mathcal{H}(G), s\in S, v\in V.$$ Define $$S\otimes_{\mathcal{H}(G)} V := (S\otimes V)/N_1.$$

On the other hand, we can view both $S$ and $V$ as $G$-representations, one acts from right, and the other from left. Let $N_2$ be the subspace of $S\otimes V$ spanned by the elements of the form $$s.g\otimes v- s\otimes g.v, \forall g\in G, s\in S, v\in V. $$ Take the quotient $$S\otimes_G V:= (S\otimes V)/N_2.$$

Now I want to show that $$S\otimes_{\mathcal{H}(G)} V= S\otimes_G V$$ by showing that $N_1=N_2$. I can show the inclusion $N_1 \subset N_2$, but the converse direction is not clear.

I'm wondering can we expect the above equality and how to prove it?

On the other hand, for any element $g\in G$, if we can find some compact open subgroup $K$, depending on $g$, such that $Kg=gK$, then one can deduce the above equality. But this seems open to me, and I don't know if one can expect this.

$\endgroup$
1
  • $\begingroup$ As @KentaSuzuki's exactly-right answer illustrates, the reason that one can pass back and forth between the (smooth) representation theory of $G$ and of $\mathcal H(G)$ is the technique of averaging over open compact subgroups, so that'll usually be the first thing you want to try for questions of this type. $\endgroup$
    – LSpice
    Commented 10 hours ago

1 Answer 1

4
$\begingroup$

I claim $N_2\subset N_1$. Indeed, for any $s\in S$, $v\in V$, and $g\in G$, there exists a compact open $K$ such that $s*e_K=s$ and $e_K*v=v$. $$\begin{align*} sg\otimes v-s\otimes gv&=(sg\otimes v-s\otimes(e_{Kg}*v))+(s\otimes(e_{Kg}*v)-s\otimes gv)\\ &=\big(s*e_{Kg}\otimes v-s\otimes (e_{Kg}*v)\big)+\big(s\otimes (e_{K}*gv)-(s*e_K)\big)\\ &\in N_1. \end{align*}$$ Conversely, I claim $N_1\subset N_2$. Indeed, for any $s\in S$, $v\in V$, and $f\in\mathcal H(G)$, we hope to show $$sf\otimes v-s\otimes fv\in N_2.$$ Lemma For any compact open $K\subset G$, we have $se_K\otimes v-s\otimes e_Kv\in N_2$.

Proof Choose a normal open subgroup $K'\subset K$ such that $s*e_{K'}=s$ and $e_{K'}*v=v$. Then $$ \begin{align*} se_{K}\otimes v-s\otimes e_Kv&=\frac1{|K/K'|}\sum_{h\in G'/G}(se_{K'h}\otimes v-s\otimes e_{hK'}v)\\ &=\frac1{|K/K'|}\sum_{h\in G'/G}(sh\otimes v-s\otimes hv)\\ &\in N_1. \end{align*}$$ QED.

For the general case assume $f=e_{KgK}$ for some $g\in G$ and compact open $K\subset G$, where $s*e_K=s$ and $e_K*v=v$. Then $$\begin{align*} se_{KgK}\otimes v-s\otimes e_{KgK}v&=(s*e_{gK})\otimes v-s\otimes (e_{Kg}*v)\\ &=(sg*e_{K})\otimes v-s\otimes (e_{K}*gv)\\ &\in N_1. \end{align*}$$

$\endgroup$
0

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.