paper-with-me

홈 › Papers

Token Pruning in Multimodal Large Language Models: Are We Solving the Right Problem?

2025-02-17 · Zichen Wen, Yifeng Gao, Weijia Li, Conghui He, Linfeng Zhang

Multimodal large language models (MLLMs) have shown remarkable performance for cross-modal understanding and generation, yet still suffer from severe inference costs. Recently, abundant works have been proposed to solve this problem with token pruning, which identifies the redundant tokens in MLLMs and then prunes them to reduce the computation and KV storage costs, leading to significant acceleration without training. While these methods claim efficiency gains, critical questions about their fundamental design and evaluation remain unanswered: Why do many existing approaches underperform even compared to naive random token selection? Are attention-based scoring sufficient for reliably identifying redundant tokens? Is language information really helpful during token pruning? What makes a good trade-off between token importance and duplication? Are current evaluation protocols comprehensive and unbiased? The ignorance of previous research on these problems hinders the long-term development of token pruning. In this paper, we answer these questions one by one, providing insights into the design of future token pruning methods.

📄 PDF Abstract BibTeX arXiv:2502.11501

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Pruning 설명 없음

Similar Papers 제목 키워드 기반

TAMP: Token-Adaptive Layerwise Pruning in Multimodal Large Language Models

2025-04-14 · Jaewoo Lee, Keyang Xuan, Chanakya Ekbote, Sandeep Polisetty 외

Multimodal Large Language Models (MLLMs) have shown remarkable versatility in understanding diverse multimodal data and tasks. However, these capabilities come with an increased model scale. While post-training pruning r…

Diversity

CATP: Contextually Adaptive Token Pruning for Efficient and Enhanced Multimodal In-Context Learning

2025-08-11 · Yanshu Li, Jianjiang Yang, Zhennan Shen, Ligong Han 외 arxiv

Modern large vision-language models (LVLMs) convert each input image into a large set of tokens that far outnumber the text tokens. Although this improves visual perception, it also introduces severe image token redundan…

Domain Adaptation

Recoverable Compression: A Multimodal Vision Token Recovery Mechanism Guided by Text Information

2024-09-02 · Yi Chen, Jian Xu, Xu-Yao Zhang, Wen-Zhuo Liu 외

With the advancement of large-scale language modeling techniques, large multimodal models combining visual encoders with large language models have demonstrated exceptional performance in various visual tasks. Most of th…

Language ModelingLanguage ModellingLarge Language Model

Script: Graph-Structured and Query-Conditioned Semantic Token Pruning for Multimodal Large Language Models

2025-12-01 · Zhongyu Yang, Dannong Xu, Wei Pang, Yingfang Yuan arxiv

The rapid growth of visual tokens in multimodal large language models (MLLMs) leads to excessive memory consumption and inference latency, especially when handling high-resolution images and videos. Token pruning is a te…

Capturing Token Tendencies for Training-Free Token Pruning in Multimodal Large Language Models

2026-07-30 · Jie Ma, Zhike Qiu, Jie Gao, Jiayi Ji 외 arxiv

While visual token pruning is essential for efficient Multimodal Large Language Models (MLLMs), existing training-free methods suffer from a critical limitation: they rely on static, instantaneous heuristics to perform i…

Trajectory Modeling