Scale AI cuts 89 tasks from SWE-Bench Pro after audit finds gaming
Read the original at theinference.org→Original headline: "Scale AI cut 89 tasks from SWE-Bench Pro after an audit found gaming"
Coverage timeline
- Oct 9, 00:32 UTC Hacker News (AI) lead source Scale AI cut 89 tasks from SWE-Bench Pro after an audit found gaming