wassname 48e5596730 Mirror honesty/credulity selections into library data/selections/ ...

Score-ranked exports (--min-score 0 --keep-per-source 10) for both axes, 46 rows each,
6 sources. Stage B produced 0/50 strict pass for both axes (authority-calibrated gate
too tight for epistemic axes); see docs/20260705_persona_library_user_perspective.md
papercut #1. These mirror the selections shipped to steering-lite.

2026-07-05 13:44:26 +08:00

..

authority_dignity_strict22_examples.md

Add strict authority scenario selection

2026-06-30 10:52:21 +08:00

authority_dignity_strict22_summary.json

Add strict authority scenario selection

2026-06-30 10:52:21 +08:00

authority_dignity_strict22.jsonl

Add strict authority scenario selection

2026-06-30 10:52:21 +08:00

credulous_skeptical_qwen3_14b_score25_scenario_scores.csv

Mirror honesty/credulity selections into library data/selections/

2026-07-05 13:44:26 +08:00

credulous_skeptical_qwen3_14b_score25.jsonl

Mirror honesty/credulity selections into library data/selections/

2026-07-05 13:44:26 +08:00

truth_over_approval_qwen3_14b_score25_scenario_scores.csv

Mirror honesty/credulity selections into library data/selections/

2026-07-05 13:44:26 +08:00

truth_over_approval_qwen3_14b_score25.jsonl

Mirror honesty/credulity selections into library data/selections/

2026-07-05 13:44:26 +08:00