undefined | Better HN

0 pointsekropotin1mo ago0 comments

> it’s clearly not the case that these things only make use of Spanish training data when you prompt them in Spanish.

It’s not! And I’ve never said that.

Anyways, I’m not even sure what we are arguing about, as it’s 100% fact that SOTA models perform better in English, the only interesting question here how much better, is it negligible or actually makes a difference in real world use-cases.

0 comments

foldr1mo ago

It’s negligible as far as I can tell. If the LLM can “speak” the language well then you can prompt it in that language and get more or less the same results as in English.

j / k navigate · click thread line to collapse

0 comments

foldr1mo ago

It’s negligible as far as I can tell. If the LLM can “speak” the language well then you can prompt it in that language and get more or less the same results as in English.

j / k navigate · click thread line to collapse