For OpenAI perhaps? Sonnet 3.7 without extended thinking is quite strong. Swe-be... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		usaar333 9 months ago \| parent \| context \| favorite \| on: GPT-4.5 For OpenAI perhaps? Sonnet 3.7 without extended thinking is quite strong. Swe-bench scores tie o3

stavros 9 months ago [–]

How do you read those scores? I wanted to see how well 3.7 with thinking did, but I can't even read that table.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact