Vision-only encoders (especially Swin-V2) predict human semantic attribute norms about as well as large language models, and multimodal encoders improve only slightly.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Seeing What Tastes Good: Revisiting Multimodal Distributional Semantics in the Billion Parameter Era
Vision-only encoders (especially Swin-V2) predict human semantic attribute norms about as well as large language models, and multimodal encoders improve only slightly.