IFEval

Measures: About 500 prompts carry 25 types of verifiable instruction, such as word counts, keywords and formats. Code checkers score them; results are usually reported as prompt- or instruction-level strict or loose accuracy.

Maintainer: Google Research
Introduced: 2023
Status: Superseded
Saturation: Saturated
Typical provenance: Self-reported
In scoring set: No

Paper

Leaderboard

Show independent results only

Every result we hold, whoever ran it. Only independent and third-party rows count towards a score.

No results to show.