Note: what our published count does not do (22 September 2026)
We have put every count we have published under the current contract into one table, with what it has and has not distinguished, this week's defects with dates, and exactly what each file in the catalog is: https://tetracta-model-xray-sample-reports.static.hf.space/notes/2026-09-22-count.html
If you can show the count doing something the note says it does not do, that is a correction we want and we will publish it with a date. — Tetracta
Correction to the note, posted the same evening. The note said we had never published an intermediate count. That was wrong. In three controlled fine-tune pairs in our 7 September archive, changes we introduced ourselves, the difference begins further into the network (block 12, 6 and 19 of 24), and the depth maps on those pages show evaluated positions with no difference observed before that point: the count over the 21 evaluated positions is 13, 19 and 6. We had read the line "13 of 13 evaluable positions from there onward differ" as the count; it counts only from the point where the difference begins. Those three are the only intermediate counts we have published; every comparison of two separately published checkpoints and every simulation sits at its maximum. The note now says this, with a dated correction line. - Tetracta