frmsaul | Dark Hacker News

1.	We Trained a Model on Office Work. It Also Got Better at Coding(surgehq.ai) 2 points by frmsaul 15 days ago \| 0 comments
2.	1 points by frmsaul 135 days ago \| discuss
3.	Is ProgramBench Impossible?(lesswrong.com) 1 points by frmsaul 146 days ago \| 0 comments
4.	Cyborg Evals(lesswrong.com) 2 points by frmsaul 154 days ago \| 1 comment
5.	You're gonna need a bigger benchmark, METR(lesswrong.com) 3 points by frmsaul 171 days ago \| 0 comments
6.	Colorful ASCII art with emojis(github.com) 1 points by frmsaul 9 years ago \| 0 comments