Hacker News new | past | comments | ask | show | jobs | submit login

Unfortunately these models are bad at math.

You'll be hard pressed to find any LLM that can even sum a long set of values without being incorrect.

That doesn't mean you can't build compute systems around these.

But it just doesn't come out of the box.




thx




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: