Skip to content
← Back to release · 2.1.285
Bug fixes / v2.1.285

Exclude truncated responses from eval averages

CHANGELOG · original

Fixed the /claude-api eval runner scaffold counting responses cut off at max_tokens in the score averages; they are now marked truncated and counted separately
Open official changelog ↗

Related documentation

Background on the feature; see the changelog above for this specific change.

Change details