Deciding on two or three interaction effects for mixed effects logistic regression in r

Question

I am trying to determine if I need a two- or three-way interaction effect and if my random slopes addition makes sense in this model.

The overall question is how do these