TY  - RPRT
TI  - If CLIP Could Talk: Understanding Vision-Language Model Representations Through Their Preferred Concept Descriptions
AU  - Reza Esfandiarpoor
AU  - Cristina Menghini
AU  - Stephen H. Bach
PY  - 2024
UR  - https://arxiv.org/abs/2403.16442
ID  - 2403.16442
ER  -