PyMC: Vectorization or Fundamentally Different

Question

is there a fundamental difference between these pymc variable definitions:

r = pymc.Uniform(lower=-1, max=1, size=2)
step = [pymc.Slice(var=[r])]

and

r1 = pymc.Uniform(lower=-1, max=1)
r2 = pymc.Uniform(lower=-1, max=1)
step = [pymc.Slice(var=[r1]), pymc.Slice(var=[r2]) ]

?

I need to update a code using the slice algorithm and some 21 variables were defined taking the later approach. I am unsure if adopting the first definition might lead to significant differences in the posterior distributions.