Hilbert 16(a) degree 8, q3, 2026-08-27

Grok 4.6 on the existing problems/hilbert16-degree-8 folder, starting from origin/main db98cb1 (q2 wrap merged).

Record

Re-fetched arXiv:2602.06888: still v3, 27 Jul 2026, 2,367 nonempty degree-8 T-schemes. Bound still ≥ 2,384. Rank 21 leftover finished on main.

Bound

Did not move. No q3/certs/new_schemes.json.

Rank 22 leftover thicken is 5/5, each with evals $46\cdot 2^{22}$. Ranks 23 and 24 are 6/6 and 3/3. Rank 26 is 1/1: four shards each with evals $46\cdot 2^{24}$, summing to $46\cdot 2^{26}$. Leftover ranks 22–26 are 15/15, 8,682,209,280 evaluations, novel empty. No q3/certs/new_schemes.json. Bound still ≥ 2,384. Odd collections of size 4 finished: twelve known M-schemes, no nest. The two open (19,3) nests remain.

Scope

Radius-1 thicken of the fifteen leftover census triangulations of twist-rank 22–26, and the two open (19,3) nests. The only leftover (19,3) census certificate is the rank-23 Harnack deg8/o22-p19-n03/(18v1(3)).pcom.

Replay:

cd problems/hilbert16-degree-8/compute
sh run_all.sh
sh q3/run_all.sh
python3 q3/collect.py