Search papers, labs, and topics across Lattice.
Thanks:
3
0
4
9
LLMs misrepresent intersectional identities by reducing complex opinions to single features, undermining their utility as synthetic survey respondents.
LLMs exhibit significant regional performance disparities, revealing critical gaps in their pretraining that could hinder their application in culturally diverse contexts.
LLMs struggle with native Greek, as evidenced by substantial performance gaps revealed by the new GreekMMLU benchmark, highlighting the need for better adaptation and evaluation methods.
