Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

SkipPLUS: Skip the first few layers to better explain vision transformers

Mehri, Faridoun, Fayyaz, Mohsen, Baghshah, Mahdieh Soleymani and Pilehvar, Mohammad Taher 2024. SkipPLUS: Skip the first few layers to better explain vision transformers. Presented at: 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), Seattle, WA, USA, 17-18 June 2024. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). pp. 204-215. 10.1109/cvprw63382.2024.00025

[thumbnail of SkipPLUS_v2.2.3.2.pdf]
Preview
PDF - Accepted Post-Print Version
Download (4MB) | Preview

Abstract

Despite their remarkable performance, the explainability of Vision Transformers (ViTs) remains a challenge. While forward attention-based token attribution techniques have become popular in text processing, their suitability for ViTs hasn't been extensively explored. In this paper, we compare these methods against state-of-the-art input attribution methods from the Vision literature, revealing their limitations due to improper aggregation of information across layers. To address this, we introduce two general techniques, PLUS and SkipPLUS, that can be composed with any input attribution method to more effectively aggregate information across layers while handling noisy layers. Through comprehensive and quantitative evaluations of faithfulness and human interpretability on a variety of ViT architectures and datasets, we demonstrate the effectiveness of PLUS and SkipPLUS, establishing a new state-of-the-art in white-box token attribution. We conclude with a comparative analysis highlighting the strengths and weaknesses of the best versions of all the studied methods. The code used in this paper is freely available at https://github.com/NightMachinery/SkipPLUS-CVPR-2024.

Item Type: Conference or Workshop Item (Paper)
Date Type: Published Online
Status: Published
Schools: Computer Science & Informatics
ISBN: 979-8-3503-6547-4
ISSN: 2160-7508
Date of First Compliant Deposit: 10 December 2024
Date of Acceptance: 7 April 2024
Last Modified: 10 Jan 2025 02:45
URI: https://orca.cardiff.ac.uk/id/eprint/173032

Actions (repository staff only)

Edit Item Edit Item

Downloads

Downloads per month over past year

View more statistics