Quantification and object perception in Multimodal Large Language Models deviate from human linguistic cognition

Open in new window