Across 17 image and 15 text models, generated images settle on far fewer senses of ambiguous words than generated sentences, and both fall well short of human diversity.
A primer in BERTology: What we know about how BERT works.Transactions of the Association for Computational Linguistics,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Where did the ambiguity go? Examining how multimodal models interpret polysemous words
Across 17 image and 15 text models, generated images settle on far fewer senses of ambiguous words than generated sentences, and both fall well short of human diversity.