Can LMs correctly distinguish🔎 confusing entity mentions in multiple documents? We study how current LMs perform QA task when provided ambiguous questions and a document set📚 that requires challenging entity disambiguation. Work done at @UTCompSci✨ w/ @xiye_nlp, @eunsolc
3
11
40
13K
19
Download Image
(2/6) We establish an ontology of five types of LM generated answers💬 (complete, partial, no, ambiguous, merged), in response to questions containing ambiguous entities. LM should generate a complete answer, pairing each disambiguated entity with each answer.
@yoonsang_ @UTCompSci @xiye_nlp @eunsolc very impressive work 👏