Fetching the paper…

Which Tokens to Use? Investigating Token Reduction in Vision Transformers · Around