(@sumaro76.bsky.social)

LLMs are increasingly used to evaluate, summarize, and even judge information. The common assumption is that, as long as their outputs align with human judgments, they can safely take on epistemic roles. In this paper, authors claim why this assumption is misleading. osf.io/preprints/ps...

loading . . .

OSF Large language models (LLMs) are widely described as artificial intelligence, yet their epistemic profile diverges sharply from human cognition. Here we show that the apparent alignment between human... https://osf.io/preprints/psyarxiv/c5gh8_v1

4 months ago