English

DAVE: Distribution-aware Attribution via ViT Gradient Decomposition

Computer Vision and Pattern Recognition 2026-02-09 v1 Artificial Intelligence Human-Computer Interaction Machine Learning

Abstract

Vision Transformers (ViTs) have become a dominant architecture in computer vision, yet producing stable and high-resolution attribution maps for these models remains challenging. Architectural components such as patch embeddings and attention routing often introduce structured artifacts in pixel-level explanations, causing many existing methods to rely on coarse patch-level attributions. We introduce DAVE \textit{(\underline{D}istribution-aware \underline{A}ttribution via \underline{V}iT Gradient D\underline{E}composition)}, a mathematically grounded attribution method for ViTs based on a structured decomposition of the input gradient. By exploiting architectural properties of ViTs, DAVE isolates locally equivariant and stable components of the effective input--output mapping. It separates these from architecture-induced artifacts and other sources of instability.

Keywords

Cite

@article{arxiv.2602.06613,
  title  = {DAVE: Distribution-aware Attribution via ViT Gradient Decomposition},
  author = {Adam Wróbel and Siddhartha Gairola and Jacek Tabor and Bernt Schiele and Bartosz Zieliński and Dawid Rymarczyk},
  journal= {arXiv preprint arXiv:2602.06613},
  year   = {2026}
}

Comments

work under review. Code will be released upon acceptance

R2 v1 2026-07-01T10:24:12.601Z