Merging does not always happen at the lowest or latest deep, it can (mostly) happen after the deep, on the way up, or (sometimes) before, on the way down, or it can similarly happen at a former deep. That is why the detection of the merge is (usually) attempted at the first term that reaches 10, 12, 15, 20, 30, 50, 80 digits. I wrote a (slow) pari script some time ago that says (after half hour of running single core) that the lowest sequence that reaches 1021845600 (the first 10 digit term after your sequence's lowest/latest deep), without going more than 2 digits above it (yes, my script is parameterized and it takes the testing a bit further), is sequence [B]3741648[/B].
Then 5455488 is only the second. (and 5924400 the third).
Also, at the former deeps, the first sequences that reaches 105229560072 (the first 12 digit term after the pre-former deep) and respective 131273006508 (the first 12 digit term after the former deep) are higher than that.
If there is other (lower than 3M) merger, then that must reach higher than 14 digits before merging. I didn't check higher, because such stuff, if it exists, even if it is difficult to detect, it will be picked by the blue page in a week or two when the updating cycle runs all around.
|
