Lune

ACL Findings2026

CrisPrune: Combining Contextual Relevance and Intrinsic Saliency for Efficient Visual Token Pruning in MLLMs

Ziniu Liu, Shuheng Zhou, Mingqing Liu, Hao Deng, Huijia Zhu

Abstract

Visual token pruning has emerged as a pivotal strategy to alleviate the computational bottleneck in Multimodal Large Language Models (MLLMs), yet it frequently compromises the integrity of visual understanding in pursuit of efficiency. Existing methods face a fundamental tension: vision-centric approaches are susceptible to the attention sink phenomenon and operate in a query-agnostic manner, whereas text-guided methods often create an overly narrow focus, discarding essential background context and failing on ambiguous queries. In this paper, we propose CrisPrune, a training-free and model-agnostic method that reconciles efficiency with understanding by integrating visual saliency and text relevance. Specifically, we introduce intrinsic visual saliency with robust normalization to identify information-rich regions characterized by significant visual features. Simultaneously, we design dual-source text relevance to synergize explicit instruction alignment with implicit scene priors. Finally, we reformulate the selection process using a Determinantal Point Process (DPP) to balance token quality and spatial diversity. Extensive experiments demonstrate that CrisPrune significantly outperforms state-of-the-art methods. On LLaVA-NeXT, it achieves a 13× decrease in FLOPs while maintaining 97% of the original performance with 94.4% of visual tokens pruned, effectively bridging the gap between efficiency and holistic understanding.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
  • Visual Instruction Tuning cites Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering
  • Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on19

  1. An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
  2. Visual Instruction TuningHaotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae LeeNeurIPS 2023 · 11,349 citations
  3. FlashAttention: Fast and Memory-Efficient Exact Attention with IO-AwarenessTri Dao, Daniel Y. Fu, Stefano Ermon, Atri Rudra et al.NeurIPS 2022 · 5,493 citations
  4. Sigmoid Loss for Language Image Pre-TrainingXiaohua Zhai, Basil Mustafa, Alexander Kolesnikov, Lucas BeyerICCV 2023 · 2,932 citations
  5. Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question AnsweringPan Lu, Swaroop Mishra, Tanglin Xia, Liang Qiu et al.NeurIPS 2022 · 2,727 citations

Related papers

  1. Beyond Attention or Similarity: Maximizing Conditional Diversity for Token Pruning in MLLMsQizhe Zhang, Mengzhen Liu, Lichen Li, Ming Lu et al.NeurIPS 2025 · 104 citations
  2. PrismPrune: Decoupling Saliency and Diversity in Attention for Efficient Visual Token Pruning in VLMsZiniu Liu, Shuheng Zhou, Mingqing Liu, Hao Deng, Huijia ZhuCVPR Findings 2026
  3. TransPrune: Token Transition Pruning for Efficient Large Vision-Language ModelAo Li, Yuxiang Duan, Jinghui Zhang, Congbo Ma et al.CVPR 2026 · 3 citations
  4. Instruction-Guided Cross-Modal Clustering for Training-Free Visual Token Pruning in Vision-Language ModelsYunqian Yu, Biao Chen, Yunya Zhang, Tonglan Xie et al.AAAI 2026
  5. Fit and Prune: Fast and Training-free Visual Token Pruning for Multi-modal Large Language ModelsWeihao Ye, Qiong Wu, Wenhao Lin, Yiyi ZhouAAAI 2025 · 99 citations
Dusk over the sea between two cliffs drawn in fine vertical lines