Balanced Token Pruning: Accelerating Vision Language Models Beyond Local Optimization

TL;DR

This paper introduces Balanced Token Pruning, a method that improves the efficiency of vision-language models by considering both local and global impacts of token removal, achieving high compression with minimal performance loss.

Contribution

We propose a novel pruning approach that balances local and global impacts, using a staged process guided by a small calibration set, to enhance token pruning effectiveness in LVLMs.

Findings

Achieves 78% compression rate with 96.7% performance retention.

Effective across various LVLMs and benchmarks.

Outperforms existing token pruning methods.

Abstract

Large Vision-Language Models (LVLMs) have shown impressive performance across multi-modal tasks by encoding images into thousands of tokens. However, the large number of image tokens results in significant computational overhead, and the use of dynamic high-resolution inputs further increases this burden. Previous approaches have attempted to reduce the number of image tokens through token pruning, typically by selecting tokens based on attention scores or image token diversity. Through empirical studies, we observe that existing methods often overlook the joint impact of pruning on both the current layer's output (local) and the outputs of subsequent layers (global), leading to suboptimal pruning decisions. To address this challenge, we propose Balanced Token Pruning (BTP), a plug-and-play method for pruning vision tokens. Specifically, our method utilizes a small calibration set to divide the pruning process into multiple stages. In the early stages, our method emphasizes the impact of pruning on subsequent layers, whereas in the deeper stages, the focus shifts toward preserving the consistency of local outputs. Extensive experiments across various LVLMs demonstrate the broad effectiveness of our approach on multiple benchmarks. Our method achieves a 78% compression rate while preserving 96.7% of the original models' performance on average. Our code is available at https://github.com/EmbodiedCity/NeurIPS2025-Balanced-Token-Pruning.