%I #18 Apr 06 2022 08:49:51
%S 2,8,4,8,16,24,32,64,48,112,64,96,128,160,192,64,512,128,64,64,128,64,
%T 256,384,64,320,256,320,384,768,384,256,384,128,896,2048,512,128,128,
%U 1280,2048,256,256,1280,3968,3840,2304,18176,3072,27136,28160,2560,8704,1024,12800,3072,6144,2560,7680,4608
%N First differences of A230624.
%C This sequence could certainly be divided by 2, just as we divided A230624 itself by 2 to get A349821. But there is a reason for not dividing this by 2: it appears that, for any power of 2, from a certain point on, the sequence is divisible by that power of 2. At present this only a conjecture. But it may provide a clue to the structure of this sequence and therefore of A230624.
%C For example, after 14 terms, the present sequence (as far as it is presently known) can be divided by 64, giving 3, 1, 8, 2, 1, 1, 2, 1, 4, 6, 1, 5, 4, 5, 6, 12, 6, 4, 6, 2, 14, 32, 8, 2, 2, 20, 32, 4, 4, 20, 62, 60, 36, 284, 48, 424, 440, 40, 136, 16, 200, 48, 96, 40, ..., which in turn can be divided by 2 after a further 14 terms.
%C So there is at least some structure here.
%H N. J. A. Sloane, <a href="/A349823/b349823.txt">Table of n, a(n) for n = 1..546</a> (based on the b-file for A230624).
%Y Cf. A230624, A349821.
%K nonn
%O 1,1
%A _N. J. A. Sloane_, Dec 31 2021