Year-end appeal: Please make a donation to the OEIS Foundation to support ongoing development and maintenance of the OEIS. We are now in our 61st year, we have over 378,000 sequences, and we’ve reached 11,000 citations (which often say “discovered thanks to the OEIS”).
%I #26 Oct 18 2024 17:59:22
%S 4681,15841,42799,52633,220729,647089,951481,1082401,1145257,1969417,
%T 2215441,3567481,4835209,5049001,5681809,6140161,6334351,8725753,
%U 10712857,11777599,12327121,13500313,14709241,22564081,22849481,22953673,23828017,27271151,28758601,30576151
%N Composite numbers k such that 2^(2^(k-1)-1) == 1 (mod k^2).
%C If 2^(k-1) == 1 (mod k) and 2^(2^(k-1)-1) == 1 (mod k), then 2^(2^(k-1)-1) == 1 (mod k^2). In fact, all such pseudoprimes are strong pseudoprimes to base 2.
%C Other terms; 951481 = 271*3511, 2215441 = 631*3511, 28758601 = 8191*3511, ... are not Fermat pseudoprimes to base 2, where 3511 is a Wieferich prime. The Wieferich prime 1093 cannot be a factor of these numbers (see A374953).
%t q[k_] := Module[{m = MultiplicativeOrder[2, k^2]}, PowerMod[2, k-1, m] == 1]; Select[Range[1, 10^6, 2], CompositeQ[#] && q[#] &] (* _Amiram Eldar_, Sep 17 2024 *)
%o (PARI) is(k) = (k > 1) && k % 2 && !isprime(k) && Mod(2, znorder(Mod(2, k^2)))^(k-1) == 1; \\ _Amiram Eldar_, Sep 17 2024
%Y Cf. A001220, A001262, A001567, A188465, A374841, A374953.
%K nonn
%O 1,1
%A _Thomas Ordowski_, Sep 17 2024
%E More terms from _Amiram Eldar_, Sep 17 2024