2 ms·
For OpenAI perhaps? Sonnet 3.7 without extended thinking is quite strong. Swe-bench scores tie o3
by usaar333 2y ago
For OpenAI perhaps? Sonnet 3.7 without extended thinking is quite strong. Swe-bench scores tie o3
- stavros 2y agoHow do you read those scores? I wanted to see how well 3.7 with thinking did, but I can't even read that table.