1 paper
Yanlin Chen, Tang Li, Xi Peng
Vision transformers (ViTs) have achieved remarkable generalization across visual domains, yet little is known about how they internally represent the structure of the world. To add…