Research map / Architectures and efficiency / Attention and transformers
A Simple Romance Between Multi-Exit Vision Transformer and Token Reduction
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
The paper studies combining multi-exit Vision Transformers with token reduction to improve ViT efficiency; only the abstract was available, so details are limited.
Problem
Vision Transformers have high computational cost, and token reduction methods need a scoring metric to decide which tokens to discard during forward propagation.
Approach
Connects multi-exit ViT architectures with token reduction, so that the scoring of unimportant tokens is informed by intermediate exits; the abstract does not state the exact mechanism.
Closest papers by meaning
- Token Transformation Matters: Towards Faithful Post-hoc Explanation for Vision Transformer (CVPR 2024 · similarity 0.89)
- Multi-criteria Token Fusion with One-step-ahead Attention for Efficient Vision Transformers (CVPR 2024 · similarity 0.89)
- Token Statistics Transformer: Linear-Time Attention via Variational Rate Reduction (ICLR 2025 · similarity 0.89)
- Frequency-Aware Token Reduction for Efficient Vision Transformer (NeurIPS 2025 · similarity 0.89)
- ImagePiece: Content-aware Re-tokenization for Efficient Image Recognition (AAAI 2025 · similarity 0.88)
- Rethinking Token Reduction with Parameter-Efficient Fine-Tuning in ViT for Pixel-Level Tasks (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.