Just a silly question, but shall we include $\sqrt{-g}$ inside the Lagrangian density in GR? In other words, is it
$$S = \int{\mathcal{L}d^4x} \Longrightarrow \mathcal{L} = \sqrt{-g}\left(\frac{1}{2\kappa}R + \frac{\Lambda}{\kappa} + \mathcal{L}_m\right)~?\tag{1}$$
Or the following
$$S = \int{\sqrt{-g}\mathcal{L}d^4x} \Longrightarrow \mathcal{L} = \frac{1}{2\kappa}R + \frac{\Lambda}{\kappa} + \mathcal{L}_m~?\tag{2}$$
Most of the actions I've seen out there (like this one) study classical particles moving in curved spacetime rather than the Einstein-Hilbert Lagrangian density itself.