paper-with-me

Papers

Similarity-Aware Token Pruning: Your VLM but Faster

2025-03-14 · Ahmadreza Jeddi, Negin Baghbanzadeh, Elham Dolatabadi, Babak Taati

The computational demands of Vision Transformers (ViTs) and Vision-Language Models (VLMs) remain a significant challenge due to the quadratic complexity of self-attention. While token pruning offers a promising solution, existing methods often introduce training overhead or fail to adapt dynamically across layers. We present SAINT, a training-free token pruning framework that leverages token similarity and a graph-based formulation to dynamically optimize pruning rates and redundancy thresholds. Through systematic analysis, we identify a universal three-stage token evolution process (aligner-explorer-aggregator) in transformers, enabling aggressive pruning in early stages without sacrificing critical information. For ViTs, SAINT doubles the throughput of ViT-H/14 at 224px with only 0.6% accuracy loss on ImageNet-1K, surpassing the closest competitor by 0.8%. For VLMs, we apply SAINT in three modes: ViT-only, LLM-only, and hybrid. SAINT reduces LLaVA-13B's tokens by 75%, achieving latency comparable to LLaVA-7B with less than 1% performance loss across benchmarks. Our work establishes a unified, practical framework for efficient inference in ViTs and VLMs.

📄 PDF Abstract BibTeX arXiv:2503.11549

Code (1)

ArmenJeddi/saint 공식 구현 jax

Methods 이 논문이 사용한 방법론

Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Feedforward Network A Feedforward Network, or a Multilayer Perceptron (MLP), is a neural network with solely densely connected layers. This is the classic neural network architecture of the…
CutMix CutMix is an image data augmentation strategy. Instead of simply removing pixels as in Cutout, we replace the removed regions with…
Mixup Mixup is a data augmentation technique that generates a weighted combination of random image pairs from the training data. Given two images and their ground truth labels:…
SAINT SAINT is a hybrid deep learning approach to solving tabular data problems. SAINT performs attention over both rows and columns, and it includes an enhanced embedding method.…
Pruning 설명 없음

Similar Papers 제목 키워드 기반

Decoupled Similarity for Task-Aware Token Pruning in Large Vision-Language Models

2026-04-13 · Kexin Ma, Jing Xiao, Chaofeng Chen, Geyong Min 외 arxiv

Token pruning has emerged as an effective approach to reduce the substantial computational overhead of Large Vision-Language Models (LVLMs) by discarding less informative visual tokens while preserving performance. Howev…

Multi-Scale And Token Mergence: Make Your ViT More Efficient

2023-06-08 · Zhe Bian, Zhe Wang, Wenqiang Han, Kangping Wang

Since its inception, Vision Transformer (ViT) has emerged as a prevalent model in the computer vision domain. Nonetheless, the multi-head self-attention (MHSA) mechanism in ViT is computationally expensive due to its cal…

Leveraging KV Similarity for Online Structured Pruning in LLMs

2025-12-08 · Jungmin Lee, Gwangeun Byeon, Yulhwa Kim, Seokin Hong arxiv

Pruning has emerged as a promising direction for accelerating large language model (LLM) inference, yet existing approaches often suffer from instability because they rely on offline calibration data that may not general…

RAPID: Layer-Wise Redundancy-Aware Pruning and Importance-Driven Token Merging for Efficient ViT

2026-06-06 · Kyumin Choi, Ikbeom Jang arxiv

Vision Transformers (ViTs) achieve strong performance but suffer from high computational costs due to quadratic self-attention complexity. Although token reduction techniques such as pruning and merging mitigate this, th…

OptiPrune: Boosting Prompt-Image Consistency with Attention-Guided Noise and Dynamic Token Selection

2025-07-01 · Ziji Lu arxiv

Text-to-image diffusion models often struggle to achieve accurate semantic alignment between generated images and text prompts while maintaining efficiency for deployment on resource-constrained hardware. Existing approa…