Overview

Visual token pruning studies how to remove or merge image and video patches before or inside vision transformers and vision-language models, reducing computation and memory for tasks such as visual question answering, video understanding, document reading, and robot control. Methods use attention or text-query relevance, feature diversity and spatial coverage, and signals such as entropy, spectral structure, or geometry; learned selectors and adaptive budgets choose different tokens or compression rates for different inputs. Pruning can happen before visual encoding, across language-model layers, or during video processing, and is evaluated on both task benchmarks and practical measures such as prefill latency, FLOPs, and memory. The central tension is that a token that seems redundant early may become important for later reasoning, while selection overhead and hardware execution can prevent fewer tokens from yielding faster end-to-end inference.

Latest papers

CardsList

298 papers23 papers in the last four weeks, down 48% on the four weeks before. 0.2% of all new papers.

  1. ASAP: Attention Sink Anchored Pruning

    May 21, 2026Jaehyuk Lee, Hanyoung Kim, Yanggee Kim +1Self-AttentionEfficient ViTs

  2. ST-SimDiff: Balancing Spatiotemporal Similarity and Difference for Efficient Video Understanding with MLLMs

    May 21, 2026Bingjun Luo, Tony Wang, Chaoqi Chen +1Multimodal Token CompressionVideo Token Compression

  3. Rethinking Token Reduction for Diffusion Models via Output-Similarity-Awareness

    May 21, 2026Hangyeol Lee, Hyojeong Lee, Joo-Young KimDiffusion TransformerDiffusion-Based Image Generation

  4. Focus-then-Context: Subject-Centric Progressive Visual Token Reduction for Vision-Language Models

    May 20, 2026Yulin Zhao, Zheng ZhangVision-Language ModelsEfficient VLM Inference

  5. Faster or Stronger: Towards Flexible Visual Place Recognition via Weighted Aggregation and Token Pruning

    May 19, 2026Zichao Zeng, June Moh Goo, Junwei Zheng +4Visual Place RecognitionEfficient ViTs

  6. EventPrune: Cascaded Event-Assisted Token Pruning for Efficient First-Person Dynamic Spatial Reasoning

    May 19, 2026Pengtao Ma, Ziliang Zhou, Ciyu Ruan +7Visual Spatial ReasoningEfficient VLM Inference

  7. Temporal Aware Pruning for Efficient Diffusion-based Video Generation

    May 18, 2026Sheng Li, Yang Sui, Junhao Ran +3Video Diffusion ModelsToken Pruning

  8. FastOCR: Dynamic Visual Fixation via KV Cache Pruning for Efficient Document Parsing

    May 17, 2026Zihan Tang, Leqi Shen, Hui Chen +7Visual AttentionEfficient VLM Inference

  9. LRCP: Low-Rank Compressibility Guided Visual Token Pruning for Efficient LVLMs

    May 15, 2026Hongyu Lu, Feng Zhang, Wenwei Jin +5Efficient VLM InferenceLarge Vision-Language Models

  10. GRIP-VLM: Group-Relative Importance Pruning for Efficient Vision-Language Models

    May 13, 2026Mingzhe Huang, Weijun Wang, Xin Ding +7Efficient VLM InferenceVisual Token Pruning

  11. CLIP Tricks You: Training-free Token Pruning for Efficient Pixel Grounding in Large VIsion-Language Models

    May 13, 2026Sangin Lee, Yukyung ChoiEfficient VLM InferenceVision-Language Grounding

  12. Keep What Audio Cannot Say: Context-Preserving Token Pruning for Omni-LLMs

    May 12, 2026Chaeyoung Jung, Kyeongha Rho, Joon Son ChungAudio-Visual UnderstandingEfficient Multimodal Inference

  13. LDDR: Linear-DPP Dynamic-Resolution Frame Sampling for Video MLLMs

    May 12, 2026Jingfeng Chen, Jiawen Qian, Wendi Deng +5Video Frame SelectionMultimodal Large Language Models

  14. ReVision: Scaling Computer-Use Agents via Temporal Visual Redundancy Reduction

    May 11, 2026Amirhossein Abaskohi, Yuhang He, Peter West +3Computer-Use AgentsLong-Horizon Agent Tasks

  15. Mutual Enhancement Between Global Tokens and Patch Tokens: From Theory to Practice

    May 11, 2026Xiusheng Huang, Xin Jiang, Jun Zhao +2Learned Image CompressionImage Tokenization

  16. EchoPrune: Interpreting Redundancy as Temporal Echoes for Efficient VideoLLMs

    May 11, 2026Jiameng Li, Minye Wu, Jiezhang Cao +2Efficient VLM InferenceToken Pruning

  17. ERASE: Eliminating Redundant Visual Tokens via Adaptive Two-Stage Token Pruning

    May 11, 2026Yuna Lee, Kyoungho Min, Yulhwa KimEfficient VLM InferenceLarge Vision-Language Models

  18. How Many Visual Tokens Do Multimodal Language Models Need? Scaling Visual Token Pruning with F^3A

    May 9, 2026YiJie Huang, Yiqun Zhang, Zhuoyue Jia +7Efficient VLM InferenceLarge Vision-Language Models

  19. PrePARE: Pre-AA Token Pruning for Frozen Multi-View Geometry Transformers

    May 8, 2026Haotang Li, Zhenyu Qi, Shaohan Henry Wang +5Multi-View GeometryEfficient ViTs

  20. An Efficient Token Compression Framework for Visual Object Tracking

    May 8, 2026Weijing Wu, Qihua Liang, Bineng Zhong +3Visual Object TrackingEfficient ViTs

  21. LookWhen? Fast Video Recognition by Learning When, Where, and What to Compute

    May 7, 2026Ali Salamatian, Anthony Fuller, Pritam Sarkar +3Efficient ViTsVideo Action Recognition

  22. Spark3R: Asymmetric Token Reduction Makes Fast Feed-Forward 3D Reconstruction

    May 7, 2026Zecheng Tang, Jiaye Fu, Qiankun Gao +53D ReconstructionEfficient ViTs

  23. LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention

    May 6, 2026Shitong Shao, Zikai Zhou, Haopeng Li +4Video EditingIn-Context Learning

  24. RTPrune: Reading-Twice Inspired Token Pruning for Efficient DeepSeek-OCR Inference

    May 1, 2026Ben Wan, Yan Feng, Zihan Tang +4Efficient VLM InferenceOptical Character Recognition

  25. YOSE: You Only Select Essential Tokens for Efficient DiT-based Video Object Removal

    Apr 30, 2026Chenyang Wu, Lina Lei, Fan Li +6Diffusion TransformerVideo Editing