“Interpretable Vision Transformers in Image Classification via SVDA” extends the group’s SVD-inspired attention work to Vision Transformers.
The paper adapts SVD-Inspired Attention to ViT image classification, using spectral and geometric indicators to inspect attention structure during training and evaluation. It strengthens the line of work around explainable AI, structured attention and interpretable computer vision.
- Authors: Vasileios Arampatzakis, George Pavlidis, Nikolaos Mitianoudis, Nikos Papamarkos
- Journal: IEEE Access, 2026
- DOI: 10.1109/ACCESS.2026.3692081
- arXiv: arXiv:2602.10994
You may also like
-
TMLR 2026: prescriptive SVD-inspired attention via spectral energy retention
-
Towards Heritage World Models: from digital twins to predictive heritage systems
-
ICASSP 2026: representation-diverse self-supervision for bioacoustic learning
-
AISTATS 2026: process-tensor tomography of SGD and training memory
-
ICPRAM 2026: two-stage angular alignment for positive-unlabeled learning