Hacker News4d ago1 min read

Why SWE-bench Verified no longer measures frontier coding capabilities

Discussed on Hacker News with 8 points and 1 comments.

Get the 10 best reads every Sunday

Curated by AI, voted by readers. Free forever.

Loading comments…

Hacker News34d ago

If you don't opt out by Apr 24 GitHub will train o

This is where you can opt out. It's absurd that they are automatically opting users into this.https://github.com/settings/copilot/features

274

Hacker News14d ago

Discussed on Hacker News with 1217 points and 504 comments.

1217

Hacker News14d ago

Discussed on Hacker News with 964 points and 512 comments.

964

Liked this? Start your own feed.