Mastodon Feed
Boosted by soatok@furry.engineer ("Soatok Dreamseeker"):
djsundog@fedi.reclaim.technology ("DJ Sundog from the *new* toot-lab") wrote:
headline: “LLMs believe false statements even after explicit warnings they’re false”
pal. no.
they believe nothing. they have not the capacity for belief.
they do not understand truth and false on any level that would matter to a non-mathematician, and even that they manage to do poorly despite decades of mostly successful usage of the concepts of true and false in computer programming.
not believing known false statements is something you would expect from an intelligence with understanding. large language models understand exactly one thing: how to model language that looks like language it has been shown. that’s all. just that. attributing it with any other understanding whatsoever demonstrates a lack of understanding on your part.
hope that fucking helps.