Files
persona-steering-template-l…/scripts
wassname 8ee98a0126 Per-axis persona jsonl + Stage B launcher for honesty/credulity
Stage A (bounded_thinking judge, 120 pairs, 0 errors, 0 judge-did-not-commit):
- truth_over_approval winner: 'Speak with the priorities of someone {persona}.' (axisD=3.00)
- credulous_skeptical winner: 'Judge the case as someone {persona} would.' (33.3% strict pass)

Stage B launched for both axes (n=50, 6 sources). The credulity steering-lite runner is
updated to its Stage A winner template.
2026-07-05 13:19:06 +08:00
..