Quote:
Originally Posted by Unregistered
I believe that I remember a theorem having to do with Generalized Mersenne primes of the form: (a^n+b^n)/(a+b).
The theorem states that if (a^n+b^n)/(a+b) is prime, then n must also be prime.
Does anyone know a reference that has a proof of this?
Thanks.

Code:
(15:18)>for(a=1,100,
for(b=a,100,
for(n=1,100,
if(((a^n+b^n)/(a+b))%1.==0 &&isprime((a^n+b^n)/(a+b)) && !isprime(n),
print(((a^n+b^n)/(a+b))","a","b","n);)
)
)
)
Quote:
14321,6,26,4
280097,12,69,4
4481,14,18,4
134417,24,57,4

shows that n=4 has counterexamples.