mlsanityData to 14 Oct 2026

PruneVid: Visual Token Pruning for Efficient Video Large Language Models

arXiv 2412.16117 ↗First posted 20 Dec 2024cs.CVcited by 67

On the map: Efficient LLM inference and serving → Visual token pruning in multimodal models

Abstract (arXiv)

In this paper, we introduce PruneVid, a visual token pruning method designed to enhance the efficiency of multi-modal video understanding. Large Language Models (LLMs) have shown promising performance in video tasks due to their extended capabilities in comprehending visual modalities. However, the substantial redundancy in video data presents significant computational challenges for LLMs. To address this issue, we introduce a training-free method that 1) minimizes video redundancy by merging spatial-temporal tokens, and 2) leverages LLMs' reasoning capabilities to selectively prune visual features relevant to question tokens, enhancing model efficiency. We validate our method across multiple video benchmarks, which demonstrate that PruneVid can prune over 80% of tokens while maintaining competitive performance combined with different model networks. This highlights its superior effectiveness and efficiency compared to existing pruning methods. Code: https://github.com/Visual-AI/PruneVid.

Key facts. Cited by 67 papers in the map (data through 14 Oct 2026). Placed in Visual token pruning in multimodal models, one of 12 directions in Efficient LLM inference and serving. Builds on 19 of its 38 references that are in the map. First posted on arXiv 20 Dec 2024.

Builds on

19 of 38 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024LLaVA-PruMerge: Adaptive Token Reduction for Efficient Large Multimodal Models22 Mar

Other references · 18

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2023GPT-4 Technical Reportcited by 18,396
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  4. 2023Visual Instruction Tuningcited by 7,869
  5. 2024LLaVA-OneVision: Easy Visual Task Transfercited by 2,719
  6. 2024Qwen2 Technical Reportcited by 2,120
  7. 2024Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysiscited by 1,413
  8. 2022Token Merging: Your ViT But Fastercited by 1,021
  9. 2021DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsificationcited by 864
  10. 2023VideoChat: Chat-Centric Video Understandingcited by 857
  11. 2023H2O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Modelscited by 832
  12. 2023EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understandingcited by 680
6 more
  1. 2023Scissorhands: Exploiting the Persistence of Importance Hypothesis for LLM KV Cache Compression at Test Timecited by 439
  2. 2021ActionCLIP: A New Paradigm for Video Action Recognitioncited by 320
  3. 2024InternVideo2: Scaling Foundation Models for Multimodal Video Understandingcited by 317
  4. 2024Tarsier: Recipes for Training and Evaluating Large Video Description Modelscited by 141
  5. 2024SlowFast-LLaVA: A Strong Training-Free Baseline for Video Large Language Modelscited by 125
  6. 2024FreeVA: Offline MLLM as Training-Free Video Assistantcited by 29

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 19; 2026: 48.

67 papers.

Sort:

By area:

  1. 2025StreamVLN: Streaming Vision-and-Language Navigation via SlowFast Context Modelingcited by 118
  2. 2025HoliTom: Holistic Token Merging for Fast Video Large Language Modelscited by 58
  3. 2025FastVID: Dynamic Density Pruning for Fast Video Large Language Modelscited by 52
  4. 2025Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodalitycited by 50
  5. 2025StreamingTOM: Streaming Token Compression for Efficient Video Understandingcited by 43
  6. 2025Video-MMLU: A Massive Multi-Discipline Lecture Understanding Benchmarkcited by 37
  7. 2026FluxMem: Adaptive Hierarchical Memory for Streaming Video Understandingcited by 28
  8. 2025LLaVA-Scissor: Token Compression with Semantic Connected Components for Video LLMscited by 26
  9. 2025A Survey of Token Compression for Efficient Multimodal Large Language Modelscited by 25
  10. 2026LVOmniBench: Pioneering Long Audio-Video Understanding Evaluation for Omnimodal LLMscited by 24
  11. 2025OmniZip: Audio-Guided Dynamic Token Compression for Fast Omnimodal Large Language Modelscited by 22
  12. 2026FlashVID: Efficient Video Large Language Models via Training-free Tree-based Spatiotemporal Token Mergingcited by 18
38 more
  1. 2025Less Is More, but Where? Dynamic Token Compression via LLM-Guided Keyframe Priorcited by 15
  2. 2025Plug-and-Play 1.x-Bit KV Cache Quantization for Video Large Language Modelscited by 13
  3. 2025MVP: Multiple View Prediction Improves GUI Groundingcited by 12
  4. 2026Attend Before Attention: Efficient and Scalable Video Understanding via Autoregressive Gazingcited by 9
  5. 2025Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Modelscited by 8
  6. 2025UniComp: Rethinking Video Compression Through Informational Uniquenesscited by 7
  7. 2025FLoC: Facility Location-Based Efficient Visual Token Compression for Long Video Understandingcited by 6
  8. 2026Think-as-You-See: Streaming Chain-of-Thought Reasoning for Large Vision-Language Modelscited by 6
  9. 2026DASH: Dynamic Audio-Driven Semantic Chunking for Efficient Omnimodal Token Compressioncited by 6
  10. 2026Unified Spatiotemporal Token Compression for Video-LLMs at Ultra-Low Retentioncited by 6
  11. 2025SpecVLM: Fast Speculative Decoding in Vision-Language Modelscited by 4
  12. 2026ForestPrune: High-ratio Visual Token Compression for Video Multimodal Large Language Models via Spatial-Temporal Forest Modelingcited by 4
  13. 2026Geometry-Guided 3D Visual Token Pruning for Video-Language Modelscited by 4
  14. 2026DocPrune:Efficient Document Question Answering via Background, Question, and Comprehension-aware Token Pruningcited by 4
  15. 2026EarlyTom: Early Token Compression Completes Fast Video Understandingcited by 4
  16. 2026AVIS: Adaptive Test-Time Scaling for Vision-Language Modelscited by 4
  17. 2026WeaveTime: Stream from Earlier Frames into Emergent Memory in VideoLLMscited by 3
  18. 2026Keeping the Evidence Chain: Semantic Evidence Allocation for Training-Free Token Pruning in Video Temporal Groundingcited by 3
  19. 2026Video-Based Reward Modeling for Computer-Use Agentscited by 3
  20. 2026OmniRefine: Alignment-Aware Cooperative Compression for Efficient Omnimodal Large Language Modelscited by 3
  21. 2026Audio-Visual Exchange-Aware Token Pruning for Efficient Audio-Visual Captioningcited by 3
  22. 2025Task-Related Token Compression in Multimodal Large Language Models from an Explainability Perspectivecited by 2
  23. 2026Compression Tells Intelligence: Visual Coding, Visual Token Technology, and the Unificationcited by 2
  24. 2026V-CAST: Video Curvature-Aware Spatio-Temporal Pruning for Efficient Video Large Language Modelscited by 2
  25. 2026LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMscited by 2
  26. 2026Temporal Aware Pruning for Efficient Diffusion-based Video Generationcited by 2
  27. 2026Swift Sampling: Selecting Temporal Surprises via Taylor Seriescited by 2
  28. 2026MemoryCard: Topic-Aware Multi-Modal Clue Compression for Long-Video Question Answeringcited by 2
  29. 2026ERA: Entropy-Guided Visual Token Pruning with Rectified Attention for Efficient MLLMscited by 2
  30. 2026OmniFocus: Query-Guided Modality-Balanced Token Compression for Omni-Modal Large Language Modelscited by 2
  31. 2025Mobile-VideoGPT: Fast and Accurate Model for Mobile Video Understandingcited by 1
  32. 2026Vision Token Reduction via Attention-Driven Self-Compression for Efficient Multimodal Large Language Modelscited by 1
  33. 2026ReDiPrune: Relevance-Diversity Pre-Projection Token Pruning for Efficient Multimodal LLMscited by 1
  34. 2026OTT-Vid: Optimal Transport Temporal Token Compression for Video Large Language Modelscited by 1
  35. 2026Rethinking RAG in Long Videos: What to Retrieve and How to Use It?cited by 1
  36. 2026OmniScope: Modality-Decoupled Token Compression for Omnimodal Large Language Modelscited by 1
  37. 2026OmniDelta: Skill-Driven Budget Allocation for Token Compression in OmniLLMscited by 1
  38. 2026Allocation Before Ranking: Decoupled Token Compression for OmniLLMscited by 1

50 most cited of 67 All 67 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026HieraVid: Hierarchical Token Pruning for Fast Video Large Language Models
  2. 2025StreamingAssistant: Efficient Visual Token Pruning for Accelerating Online Video Understanding
  3. 2026Adaptive Two-Stage Visual Token Pruning for Efficient Inference in Video-Language Models
  4. 2026Multi-Image Visual Token Pruning in Large Visual Language Models
  5. 2025Script: Graph-Structured and Query-Conditioned Semantic Token Pruning for Multimodal Large Language Models
  6. 2026FoveaTree: Information-Density Trees for Query-Guided Visual Token Pruning in Video Large Language Models
  7. 2026SinkPruner: Sink-Free Visual Token Pruning for Multimodal Large Language Models
  8. 2025Multi-Cue Adaptive Visual Token Pruning for Large Vision-Language Models