OverviewFilesCommitsIssuesDocsReleasesPeers

atomspace-layer-discovery commits / e64df84

experiments/level2: RESULTS-LV — the scorer fix came after the first scoring, not before
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
e64df8482446858f6cdab16b39cf863d03eec89c
by plankatron <fleeky@gmail.com> · 2026-10-19T02:25:22.000Z

Changed files (1) +3 -2

experiments/level2/RESULTS-LV.md +3 -2
……@@ -1,8 +1,9 @@
11# Test LV — Linas's clusterer as layer 1 — results
22
33Run `results-LV/1792366731` (code `1f8e6b6`, 10 seeds). Scored against PREREG-LV (`bd0d982`) by `summarize-LV.py`. Scorer fix
4before reading the verdicts: the L-MST reference first picked up Test A-U's MPG row (0.680); it now reads the MST
5row (0.700), as pre-registered. No verdict changed.
4after the first scoring (the verdicts had been printed): the L-MST reference first picked up Test A-U's MPG row
5(0.680); it now reads the MST row (0.700), as pre-registered. No verdict changed. (Corrected 2026-10-18: an earlier
6version of this note said the fix came before reading the verdicts; it did not.)
67
78| | prediction (30K) | verdict |
89|---|---|---|