Skip to content

Comment on “Disregard That” Attacksparent

Comments

Thanks that's useful.

So it's still one stream of tokens as far as the LLM is concerned, but there is some emphasis in training on "trust the system prompt", have I got that right?

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.