Given a discrete group $\Gamma$, in the definition of its group von Neumann algebra $L(\Gamma)$, the following map is considered, $$\lambda: \Gamma \to \mathcal{B}\left(l^2\left(\Gamma\right)\right),$$ where $\lambda(g)\delta_h:=\delta_{gh}$.
This might be a silly question. But what is $l^2(\Gamma)$? Is it the set of functions $l: \Gamma \to \mathbb{C}$ such that $\sum_{g\in\Gamma}\left|l(g)\right|^2<\infty$ (square summable)? But here should we worry about the cardinality of group $\Gamma$? Do we apply the usual definition of uncountable sum?
Also, what is the relationship between $L(\Gamma)$ and group algebra $\mathbb{C}[\Gamma]$? Or are they not related, but $l^2(\Gamma)$ and $\mathbb{C}[\Gamma]$ are related?
Thanks.
$\ell^2(\Gamma):=\{l:\Gamma\rightarrow\mathbb C \mid \sum_{g\in\Gamma}|l(g)|^2<\infty \}$ as you described.
There are a few different ways to define $\sum_{g\in\Gamma} |l(g)|$.
Frankly, it may not be worth the effort to worry about groups that are uncountable in this scenario. After all, life is too short for non-separable Hilbert spaces, and "nonseparable Hilbert spaces are in some sense artifacts and can almost always be avoided". However, it's common to consider the uncountable groups with a nondiscrete topology, such as Lie groups for the study of group von Neumann algebras, hence the last perspective is really superior in many ways.
$\mathbb C[\Gamma]$ is a subalgebra of $L(\Gamma)$, and in a precise way, $L(\Gamma)$ is a topological closure of $\mathbb C[\Gamma]$ in $\mathcal B(\ell^2(\Gamma))$, see e.g. Wikipedia.