Marius CiepluchandClaude Opus 5 66a51c674c docs(learning-harness): round 5, and what it says about the experiment itself ...

Round 5 re-ran round 1's content through the lever it should have used --
counterexamples, rendered to the detector as "Worked examples" rather than
"Known non-issues, do not report these" -- and was rejected because one pair
of eight moved, at p = 1.000.

So the experiment I designed to give the demo lever a fair test could not
give it one. Correcting my authoring error changed the reason for the verdict
and not the verdict, because the rule cannot tell a lever that helps from a
lever that does nothing.

Five rounds, $2.17, zero decisions distinguishable from chance.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01T7XD7x5KonMdmGRkp63EQG

2026-09-07 08:09:50 +02:00

..

2026-09-06-candidate-ceiling-all-slices.json

feat(constrained-detector): the ceiling, measured offline across every slice

2026-09-06 18:44:19 +02:00

2026-09-06-candidate-ceiling.json

measure: the candidate generator must be the IR, not the ruleset

2026-09-06 16:04:07 +02:00

2026-09-06-corpus-hygiene.json

feat(learning-harness): absence rows travel with their registration

2026-09-06 12:39:55 +02:00

2026-09-06-fixture-difficulty.json

feat(learning-harness): declared twins and a fixture-difficulty check

2026-09-06 13:09:56 +02:00

2026-09-06-gate-baseline.txt

test(learning-harness): the gate baseline and the degradation matrix

2026-09-06 13:20:23 +02:00

2026-09-06-learning-families.json

fix(learning-harness): the published table lost a whole slice

2026-09-06 20:44:56 +02:00

2026-09-06-mechanism-lever-split.json

feat(learning-harness): one teacher rule across every learning path

2026-09-06 10:15:44 +02:00

2026-09-06-trajectory-divergence.json

measure(detector-optimizer): the detector never runs the same investigation twice

2026-09-06 15:49:54 +02:00

2026-09-06-vibe-py-holdout-hunter-findings.json

docs(learning-harness): brief for a classified, self-improving detector set; fold closed-loop-integrity

2026-09-06 08:08:01 +02:00

2026-09-06-vibe-py-holdout-hunter-sonnet5.json

docs(learning-harness): brief for a classified, self-improving detector set; fold closed-loop-integrity

2026-09-06 08:08:01 +02:00

2026-09-07-evolve-rounds-injection.json

docs(learning-harness): round 5, and what it says about the experiment itself

2026-09-07 08:09:50 +02:00