Analyzing Vision Transformers for Image Classification in Class Embedding Space | Read Paper on Bytez