Decoding the Functional Roles of Register and High-Norm Patch Tokens in Vision Transformers
Paper recorded by Signals 4 on 2026-10-02 in cs.CV. Abstract reproduced from arXiv; link to the original below.
Published 2026-10-02 on arXiv · recorded by Signals 4 on 2026-10-05
Category: cs.CV · 计算机视觉 · first seen 2026-10-05
Abstract
Self-supervised Vision Transformers (ViTs), such as DINOv2, learn rich visual representations, but the functions of their internal tokens remain poorly understood. Recent architectures introduce dedicated register tokens to reduce high-norm out- lier patch tokens that emerge in background re- gions, yet the semantic and functional roles of both token types have not been fully established. In this
Read on arXiv →
Cite this page: Decoding the Functional Roles of Register and High-Norm Patch Tokens in Vision Transformers: the #366 most recent of 756 cs.CV papers we have recorded (as of 2026-10-02). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/decoding-the-functional-roles-of-register-and-high-norm-patch-tokens-in-vision-t.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable:
papers.json
Get 4 AI signals a day by email — free.
Get 4 AI signals a day by email — free