Interpretability-Guided Soft Pruning of Attention Heads in Vision Transformers 文章

ArXiv CS.CV2026-08-04PAPERen作者: Kamil Ksi\k{a}\.zek, Piotr Suszy\'nski, Micha{\l} Jan W{\l}odarczyk, Jacek Tabor, Przemys{\l}aw Biecek

Interpretability-Guided Soft Pruning of Attention Heads in Vision Transformers · 相关技术