373 kB
10 files
Updated 7 days ago
Name
Size
ifbench-200
ifeval-200
README.md224 Bytes
xet
SOURCES.md498 Bytes
xet
README.md

evals-public

Frozen evaluation cohorts built only from public sources, with seeds and scorer code, so results can be reproduced. Each cohort: <name>/tasks.jsonl, <name>/scorer/, <name>/SOURCES.md (origin + licence).

Total size
373 kB
Files
10
Last updated
Nov 5
Pre-warmed CDN
US EU US EU

Contributors