Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Demystifying Evals for AI Agents (anthropic.com)
2 points by vinhnx 16 days ago | hide | past | favorite | 1 comment


"You won't know if your graders are working well unless you read the transcripts and grades from many trials."

Anyone can recommend tools for reading and analysing transcripts? I assume this is regarding data that CC is writing in .claude directory.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: