Can VLMs Recall Factual Associations From Visual References?Published in EMNLP 2025, 2025Direct LinkShare on Twitter Facebook LinkedIn Previous Next