Live data from Hacker News

LLM Evals Are Just Tests. Why Are We Making This So Complicated?

cameronwestland.com

1–3 of 3 posts

Re: LLM Evals Are Just Tests. Why Are We Making This So Complicated?

#3

So, did the tests allow you to build a system that never confused existing features with new features? That seems like the problem statement, but I think I'm only seeing probabilistic testing.

Never? No. Way less likely? Yes!

In dev we do 100 consistency checks and get green. In CI we do 10.