LeGrad: An Explainability Method for Vision Transformers via Feature Formation Sensitivity
Walid Bousselham(IBM (United States)), Hilde Kuehne(University of Bonn), Sofian Chaybouti(Goethe University Frankfurt), Hendrik Strobelt(Technische Universität Dresden), Angie Boggust(Massachusetts Institute of Technology)
Cited by 6
Related Papers
Gene mention normalization and interaction extraction with context models and sentence motifs
|Genome biology|2008|63
Grounding Everything: Emerging Localization Properties in Vision-Language Transformers
|Unknown|2024|38
Efficient Self-Ensemble for Semantic Segmentation
|arXiv (Cornell University)|2021|15
Efficient Self-Ensemble Framework for Semantic Segmentation
|arXiv (Cornell University)|2021|12
Learning Situation Hyper-Graphs for Video Question Answering
|Unknown|2023|8