logoalt Hacker News

viraptoryesterday at 8:52 PM3 repliesview on HN

Like identifying names of skateboard tricks from the description? https://skatebench.t3.gg/


Replies

alargemooseyesterday at 9:23 PM

I don’t care how practical it may or may not be, this is my new favorite LLM benchmark

stevageyesterday at 9:38 PM

I couldn't find an about page or similar?

show 1 reply
hmottestadyesterday at 9:25 PM

o3-pro is better than 5.2 pro! And GPT 5 high is best. Really quite interesting.