mertyg / vision-language-models-are-bows

Experiments and data for the paper "When and why vision-language models behave like bags-of-words, and what to do about it?" Oral @ ICLR 2023
MIT License
222 stars 14 forks source link

fast #12

Closed vinid closed 1 year ago