YUKINO reports state-of-the-art scores on SugarCREPE and Winoground by injecting image-derived pseudo-tokens into captions and comparing yes versus no prompts, a protocol that likely inflates the results.
Winoground: Probing vision and language models for visio-linguistic compositionality,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.CV 1years
2024 1verdicts
REJECT 1roles
dataset 1polarities
unclear 1representative citing papers
citing papers explorer
-
A New Method to Capturing Compositional Knowledge in Linguistic Space
YUKINO reports state-of-the-art scores on SugarCREPE and Winoground by injecting image-derived pseudo-tokens into captions and comparing yes versus no prompts, a protocol that likely inflates the results.