Why SWE-bench Verified no longer measures frontier coding capabilities (openai.com) 2 points by gmays 8mo ago ↗ HN
1 comment
[ 3.1 ms ] story [ 16.9 ms ] thread