Fetching the paper…

Single-Layer Vision Transformers for More Accurate Early Exits with Less Overhead · Around