Reducing Token Redundancy in LVLMs: A Systematic Review of Token Pruning Methods

ACL ARR 2026 January Submission5444 Authors

Keywords: MLLM, pruning

Abstract: Large Vision-Language Models (LVLMs) excel at visual understanding but face severe computational bottlenecks when processing high-resolution images and long videos due to massive visual token counts. Token pruning mitigates this by selectively removing less informative tokens while maintaining performance. However, existing methods vary widely in pruning location (vision encoder vs. LLM decoder), importance criteria (attention vs. similarity vs. learned scores), and application strategy, lacking systematic comparison. This survey presents the first comprehensive review of token pruning for LVLMs. We propose a taxonomy categorizing methods into vision-side, LLM-side, and hybrid paradigms, systematically analyze token selection mechanisms and pruning strategy. We further discuss evaluation protocols and identify key challenges including prompt-adaptive pruning and hardware-aware design. Our survey provides a structured foundation for this rapidly growing research area.

Paper Type: Long

Research Area: Multimodality and Language Grounding to Vision, Robotics and Beyond

Research Area Keywords: Efficient/Low-Resource Methods for NLP, Summarization

Contribution Types: Surveys

Languages Studied: English

Submission Number: 5444