Shay b1f12e1dce chore(evals): refresh stale committed reports ... G3_numerics report.json: refusal reason-string drift (PR #359 made recognized-but-uninjectable refusals more specific); verdicts and wrong=0 identical. train_sample coverage probe: admitted_solved 4->6 reflecting reader coverage gains earlier in the window; wrong=0, safety_rail_intact. Regenerated via their runners.		2026-06-03 01:01:50 -07:00
..
G1_verb_classes/v1	feat(G1): verb-classes capability axis (ADR-0131.G.1)	2026-05-23 15:39:14 -07:00
G2_comparatives	fix(ADR-0131.G.2): rebase + mastery hardening — quarter/third fraction anchors, gate regex, boundary refusals (#196)	2026-05-23 19:28:09 -07:00
G3_numerics	chore(evals): refresh stale committed reports	2026-06-03 01:01:50 -07:00
G4_multi_clause
G5_aggregate	feat(ADR-0131.G.5): aggregate answer composition — combined/together cues wired, axis lane 20/20, wrong==0 (#197)	2026-05-23 19:42:55 -07:00
S1_rate_events	feat(ADR-0136.S.1): rate/event statement parsing — capacity + earnings shapes, axis lane 20/20, wrong==0, gsm8k-0014 admits (#201)	2026-05-23 20:36:01 -07:00
S3_compound_initial_mutation	feat(ADR-0136.S.3): compound initial-mutation extractor — one shape, gsm8k-0010 barrier shift, wrong==0 (#207)	2026-05-23 21:58:55 -07:00
S4_novel_initial_form	feat(ADR-0136.S.4): novel-initial-form parser extension + rescan v4 (#210)	2026-05-23 22:34:51 -07:00
__init__.py