wassname a80a0d56ea Add overall_score + configurable thresholds + exclude-confound-dims ...

- overall_score = axis_delta - off_axis - style - 3*(echo+refusal+nocommit)
  Lets you rank ALL scenarios by net axis signal and take top N, not just strict-pass filter.
- --axis-delta-threshold (default 3.0): lower to 2.0 for hard-to-steer axes.
- --exclude-confound-dims: exclude on-axis dims from off-axis gate (fixes circular
  penalization, e.g. honesty_truthfulness for the honesty axis).
- --off-axis-threshold (default 2.0): configurable off-axis gate.

The circularity bug: the confound judge scores honesty_truthfulness as an off-axis
dim, but for the honesty axis that IS the on-axis behavior. Excluding it from the
gate max is a correctness fix, not a relaxation.

2026-07-06 06:11:13 +08:00

..

scenario_sources

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

bounded_thinking_judge_liveproof.py

Add bounded-thinking judge: phase-1 bounded thinking + phase-2 force-answer

2026-07-05 12:53:24 +08:00

bounded_thinking_judge.py

Add bounded-thinking judge: phase-1 bounded thinking + phase-2 force-answer

2026-07-05 12:53:24 +08:00

build_hf_dataset.py

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

docs_results.py

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

export_authority_steering_selection.py

Filter selected authority scenarios by strict pass

2026-06-30 10:51:36 +08:00

export_persona_template_stats.py

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

parse_stage_a.py

Add parse_stage_a.py: rank Stage A templates by strict_pass_rate then axis_delta

2026-07-05 22:24:01 +08:00

plot_on_off_axis.py

docs: use one Quarto source for README and Pages

2026-06-25 13:06:12 +08:00

prepare_authority_steering_selection.py

Tighten pure authority selection

2026-07-01 00:29:31 +08:00

readme_plot.py

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

run_honesty_credulity_stage_b.sh

Per-axis persona jsonl + Stage B launcher for honesty/credulity

2026-07-05 13:19:06 +08:00

summarize_model_matrix.py

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

sync_template_library.py

docs: keep generated stats out of data

2026-06-13 19:12:24 +08:00

template_catalog.py

organize persona and scenario data paths

2026-06-29 19:16:41 +08:00

update_readme_model_matrix.py

repo: move scenario sources and committed results

2026-06-30 09:45:32 +08:00

update_readme_results_table.py

organize persona and scenario data paths

2026-06-29 19:16:41 +08:00

validate_persona_axes_openrouter.py

Add overall_score + configurable thresholds + exclude-confound-dims

2026-07-06 06:11:13 +08:00