Hacker News new | ask | show | jobs
by stared 10 days ago
Nice!

I remember "Baba is Eval" (https://fi-le.net/baba/), released 11 months ago, back when Claude Opus 4 was the strongest model. Back then, I was surprised how poor was it even at the first level.

I am happy to see an another approach - and indeed, with much stronger results.
1 comments
meffmadd 10 days ago
Yes that was the post that inspired me to build this.

While I did implement a more comprehensive harness with path finding tools etc. the models themselves have improved significantly.

link
stared 10 days ago
I saw you mentioned that!

Anyway: did you test it with Claude Opus 4.8?

link