wassname bcbdb9cc6f feat: multi-label moral foundation ratings with z-scored frame averaging and human calibration ...

- Add scripts/07_multilabel.py: LLM judge rates all 7 foundations per vignette
  using violation (forward) and acceptability (reverse) frames
- Foundation definitions drawn from Clifford et al. (2015) survey rubric
- Z-score each frame per foundation before averaging to cancel range bias
- Calibrate LLM Likert → human % via per-foundation OLS (classic set only)
- Add scripts/07a_merge_labels.py: merges llm_* and calibrated_* into vignette files
- Update README and HF dataset card with methodology and calibration quality table
- Classic set: 80.3% dominant-foundation accuracy, Pearson r 0.69-0.89 per foundation

2026-05-03 12:48:14 +08:00

..

01_download.py

init

2026-04-30 17:10:09 +08:00

02_rewrite.py

Implement N-token evaluation with guided rollouts

2026-05-01 21:44:14 +08:00

02a_backfill_labels.py

add human labels

2026-05-02 15:58:08 +08:00

03_eval.py

Implement N-token evaluation with guided rollouts

2026-05-01 21:44:14 +08:00

04_validate.py

decent

2026-04-30 21:22:07 +08:00

05_drift.py

decent

2026-04-30 21:22:07 +08:00

05_upload_hf.py

feat: multi-label moral foundation ratings with z-scored frame averaging and human calibration

2026-05-03 12:48:14 +08:00

06_consistency.py

fix airisk self_violate actor archetype + simplify to 2 conditions

2026-04-30 21:29:38 +08:00

07_multilabel.py

feat: multi-label moral foundation ratings with z-scored frame averaging and human calibration

2026-05-03 12:48:14 +08:00

07a_merge_labels.py

feat: multi-label moral foundation ratings with z-scored frame averaging and human calibration

2026-05-03 12:48:14 +08:00

smoke_batch_parity.py

batch the eval: guided_rollout_batch + 4× speedup

2026-05-03 06:42:17 +08:00