SWE-bench Verified no longer measures frontier coding capabilities
(openai.com)
324 points
by kmdupree
a day ago |
171 comments
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
()
