Is this sequence defined for all n?
From Jeppe Stig Nielsen, Sep 16 2015: (Start)
Numbers of this form are sometimes called extended generalized Fermat numbers.
If we restrict ourselves to the case y=1, we get instead the sequence A123599, therefore a(n) <= A123599(n) for all n. Can this be an equality for some n > 4?
The formula x^(2^m) + y^(2^m) also gives the decreasing chain {A000040, A002313, A002645, A006686, A100266, A100267, ...} of subsets of the prime numbers if we drop the requirement that x != y and take all primes (not just the smallest one) with m greater than some lower bound.
(End)
For more terms (the values of max(x,y)), see A291944.  Jeppe Stig Nielsen, Dec 28 2019
