Leveraging Cross-Modal Neighbor Representation for Improved CLIP Classification | Read Paper on Bytez