⁸Float, because the package call causes FriCAS to convert (2 + 3) and 4 to type Float. Before the sum is converted, it is given a target type (see below) of Float by FriCAS and then evaluated. The target type causes the + from Float to be used.