HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by throwaway85825 | Hacker News Reader
Full thread
throwaway85825
·
'AI' still fails the alignment grannybench.
View on HN
JacobAsmuth
·
Yup. gorgetbench, grannybench, and tebufranksalso bench are all completely unsaturated.
Reply on news.ycombinator.com