If you spend any time around Computer Vision teams or tooling today, you’ll hear a flurry of model names: ViT, SAM, CLIP, YOLO, and so...