A function to numerically stabilize an exponentiation. For internal use only
stabExp(exps)
exps
the vector to be exponentiated
the vector with the maximum subtracted