I haven’t understood the details yet, but just a basic question. Your “randomized Turing machines”, do they have a random oracle or a random advice? In other words, do you assume the lookup of F(i,j) takes time poly(logi,logj) or poly(i,j)?
It does not matter. Let’s say poly(i,j). The most it is going change is how large a gap is needed between f and g.
I haven’t understood the details yet, but just a basic question. Your “randomized Turing machines”, do they have a random oracle or a random advice? In other words, do you assume the lookup of F(i,j) takes time poly(logi,logj) or poly(i,j)?
It does not matter. Let’s say poly(i,j). The most it is going change is how large a gap is needed between f and g.