frontier-predictions-v1
·
16 commits
to main
since this release
Raw per-paragraph predictions behind every published Arabic client-frontier verdict (Paper B, RESULTS.md). Re-score any run:
pip install interscript-ml-tools[sadeed]
interscript-sadeed-eval score --preds <file> --data Misraj/SadeedDiac-25 --key student
run-006 carries the 2026-09-05 correction (RESULTS.md): its rows re-score to 5.0821, the corrected 2.0 number — the tooling that published this release is the tooling that caught it.