Marius CiepluchandClaude Fable 5.1 ad20405fc2 docs(benchmark): where the hunter's false alarms come from ...

Sweep over the pilot's hunter records: most 'vulnerable' suspicions on ordinary
pushes sit far from the diff. Scope to the change, then confidence, then the
config_secrets family; 5/12 recall kept at 8.8 false alarms per 100 (in-sample).

Co-Authored-By: Claude Fable 5.1 <[email protected]>

2026-10-13 08:35:04 +02:00

..

accepted_repos.json

feat(benchmark): freeze the valid-benchmark pilot, 12 verified positives + 240 negatives

2026-10-12 19:42:45 +02:00

catalog.toml

feat(benchmark): freeze the valid-benchmark pilot, 12 verified positives + 240 negatives

2026-10-12 19:42:45 +02:00

filter_sweep.py

docs(benchmark): where the hunter's false alarms come from

2026-10-13 08:35:04 +02:00

metrics_a_interim_163.json

docs(benchmark): pilot numbers, hunter arm final and push hook interim

2026-10-13 08:13:02 +02:00

metrics_b_filtered.json

docs(benchmark): pilot numbers, hunter arm final and push hook interim

2026-10-13 08:13:02 +02:00

metrics_b_raw.json

docs(benchmark): pilot numbers, hunter arm final and push hook interim

2026-10-13 08:13:02 +02:00

NOTES.md

docs(benchmark): where the hunter's false alarms come from

2026-10-13 08:35:04 +02:00

scored_a_interim_163.json

docs(benchmark): pilot numbers, hunter arm final and push hook interim

2026-10-13 08:13:02 +02:00

scored_b_filtered.json

docs(benchmark): pilot numbers, hunter arm final and push hook interim

2026-10-13 08:13:02 +02:00

scored_b_raw.json

docs(benchmark): pilot numbers, hunter arm final and push hook interim

2026-10-13 08:13:02 +02:00

sources.json

feat(benchmark): freeze the valid-benchmark pilot, 12 verified positives + 240 negatives

2026-10-12 19:42:45 +02:00

verification.json

feat(benchmark): freeze the valid-benchmark pilot, 12 verified positives + 240 negatives

2026-10-12 19:42:45 +02:00