Which Tokens to Use? Investigating Token Reduction in Vision Transformers

Last updated