Shareable benchmark table

Reference matches and novel-valid findings

Closed-world confusion counts beside the separate expert-validity review of findings outside the reference set.

Table 03 · Finding dispositions

Reference matches and novel-valid findings

Active researchReconstructed scoring completeFrozen2 retained non-successesOfficial judge not run
Closed-world confusion counts beside the separate expert-validity review of findings outside the reference set.
Finding dispositionGPT-5.6 Sol · HighGrok 4.6 Build · High
TP / FP / FN, closed world52 / 51 / 7747 / 56 / 82
Novel-valid findings4951
Expert-validity precision98.06%96.12%
Conventional accuracyN/AN/A
Table 03Closed-world confusion counts beside the separate expert-validity review of findings outside the reference set.

An unmatched finding is not automatically invalid. Conventional accuracy is undefined for this open-ended audit task.

BMSbms.kyokasuigetsu.xyz
Static image: https://bms.kyokasuigetsu.xyz/tables/campaign-findings.png