{"total":79,"rows":[{"id":"2406.02069","title":"PyramidKV: Dynamic KV Cache Compression based on Pyramidal Information Funneling","created":"2024-06-04","cited_by_count":432,"area":"inference"},{"id":"2410.10781","title":"When Attention Sink Emerges in Language Models: An Empirical View","created":"2024-10-14","cited_by_count":208,"area":"inference"},{"id":"2405.14366","title":"MiniCache: KV Cache Compression in Depth Dimension for Large Language Models","created":"2024-05-23","cited_by_count":125,"area":"inference"},{"id":"2411.15024","title":"DyCoke: Dynamic Compression of Tokens for Fast Video Large Language Models","created":"2024-11-22","cited_by_count":96,"area":"inference"},{"id":"2505.21334","title":"HoliTom: Holistic Token Merging for Fast Video Large Language Models","created":"2025-05-27","cited_by_count":58,"area":"inference"},{"id":"2506.15745","title":"InfiniPot-V: Memory-Constrained KV Cache Compression for Streaming Video Understanding","created":"2025-06-18","cited_by_count":58,"area":"inference"},{"id":"2505.15269","title":"LiveVLM: Efficient Online Video Understanding via Streaming-Oriented KV Cache and Retrieval","created":"2025-05-21","cited_by_count":54,"area":"inference"},{"id":"2505.18227","title":"Token Reduction Should Go Beyond Efficiency in Generative Models -- From Vision, Language to Multimodality","created":"2025-05-23","cited_by_count":50,"area":"inference"},{"id":"2512.00891","title":"Accelerating Streaming Video Large Language Models via Hierarchical Token Compression","created":"2025-11-30","cited_by_count":37,"area":"inference"},{"id":"2412.20504","title":"ReTaKe: Reducing Temporal and Knowledge Redundancy for Long Video Understanding","created":"2024-12-29","cited_by_count":33,"area":"inference"},{"id":"2507.07990","title":"Multi-Granular Spatio-Temporal Token Merging for Training-Free Acceleration of Video LLMs","created":"2025-07-10","cited_by_count":31,"area":"inference"},{"id":"2602.04804","title":"OmniSIFT: Modality-Asymmetric Token Compression for Efficient Omni-modal Large Language Models","created":"2026-02-04","cited_by_count":25,"area":"inference"},{"id":"2510.07293","title":"AudioMarathon: A Comprehensive Benchmark for Long-Context Audio Understanding and Efficiency in Audio LLMs","created":"2025-10-08","cited_by_count":22,"area":"inference"},{"id":"2511.14582","title":"OmniZip: Audio-Guided Dynamic Token Compression for Fast Omnimodal Large Language Models","created":"2025-11-18","cited_by_count":22,"area":"inference"},{"id":"2502.00791","title":"Vision-centric Token Compression in Large Language Model","created":"2025-02-02","cited_by_count":21,"area":"inference"},{"id":"2602.08024","title":"FlashVID: Efficient Video Large Language Models via Training-free Tree-based Spatiotemporal Token Merging","created":"2026-02-08","cited_by_count":18,"area":"inference"},{"id":"2512.06866","title":"Less Is More, but Where? Dynamic Token Compression via LLM-Guided Keyframe Prior","created":"2025-12-07","cited_by_count":15,"area":"inference"},{"id":"2602.06953","title":"DAWN: Dependency-Aware Fast Inference for Diffusion LLMs","created":"2026-02-06","cited_by_count":13,"area":"inference"},{"id":"2512.07580","title":"When Token Pruning is Worse than Random: Understanding Visual Token Information in VLLMs","created":"2025-12-08","cited_by_count":12,"area":"inference"},{"id":"2505.10118","title":"Why 1 + 1 < 1 in Visual Token Pruning: Beyond Naive Integration via Multi-Objective Balanced Covering","created":"2025-05-15","cited_by_count":11,"area":"inference"},{"id":"2506.20066","title":"ToSA: Token Merging with Spatial Awareness","created":"2025-06-24","cited_by_count":11,"area":"inference"},{"id":"2605.09649","title":"Make Each Token Count: Towards Improving Long-Context Performance with KV Cache Eviction","created":"2026-05-10","cited_by_count":11,"area":"inference"},{"id":"2512.12560","title":"StreamingAssistant: Efficient Visual Token Pruning for Accelerating Online Video Understanding","created":"2025-12-14","cited_by_count":10,"area":"inference"},{"id":"2602.08005","title":"ResidualKV: Residual-Based KV Cache Compression for Efficient Long-Context Inference","created":"2026-02-08","cited_by_count":9,"area":"inference"},{"id":"2509.09955","title":"Adaptive Token Merging for Efficient Transformer Semantic Communication at the Edge","created":"2025-09-12","cited_by_count":7,"area":"inference"},{"id":"2601.13143","title":"FastAV: Efficient Token Pruning for Audio-Visual Large Language Model Inference","created":"2026-01-19","cited_by_count":7,"area":"inference"},{"id":"2511.05534","title":"FlowMM: Cross-Modal Information Flow Guided KV Cache Merging for Efficient Multimodal Context Inference","created":"2025-10-29","cited_by_count":6,"area":"inference"},{"id":"2512.10324","title":"EchoingPixels: Aliasing-Resistant Joint Token Reduction for Audio-Visual LLMs","created":"2025-12-11","cited_by_count":6,"area":"inference"},{"id":"2602.21233","title":"AngelSlim: A more accessible, comprehensive, and efficient toolkit for large model compression","created":"2026-02-07","cited_by_count":6,"area":"inference"},{"id":"2603.15685","title":"DASH: Dynamic Audio-Driven Semantic Chunking for Efficient Omnimodal Token Compression","created":"2026-03-15","cited_by_count":6,"area":"inference"},{"id":"2603.21957","title":"Unified Spatiotemporal Token Compression for Video-LLMs at Ultra-Low Retention","created":"2026-03-23","cited_by_count":6,"area":"inference"},{"id":"2506.08373","title":"Draft-based Approximate Inference for LLMs","created":"2025-06-10","cited_by_count":5,"area":"inference"},{"id":"2509.12159","title":"EfficientUICoder: A Bidirectional Token Compression Framework for Efficient MLLM-Based UI Code Generation","created":"2025-09-15","cited_by_count":5,"area":"inference"},{"id":"2603.01400","title":"Token Reduction via Local and Global Contexts Optimization for Efficient Video Large Language Models","created":"2026-03-02","cited_by_count":5,"area":"inference"},{"id":"2502.06474","title":"UniMoD: Efficient Unified Multimodal Transformers with Mixture-of-Depths","created":"2025-02-10","cited_by_count":4,"area":"inference"},{"id":"2605.00392","title":"RTPrune: Reading-Twice Inspired Token Pruning for Efficient DeepSeek-OCR Inference","created":"2026-05-01","cited_by_count":4,"area":"inference"},{"id":"2605.14458","title":"OmniDrop: Layer-wise Token Pruning for Omni-modal LLMs via Query-Guidance","created":"2026-05-14","cited_by_count":4,"area":"inference"},{"id":"2605.30010","title":"EarlyTom: Early Token Compression Completes Fast Video Understanding","created":"2026-05-28","cited_by_count":4,"area":"inference"},{"id":"2504.18579","title":"Sparsity Forcing: Reinforcing Token Sparsity of MLLMs","created":"2025-04-23","cited_by_count":3,"area":"inference"},{"id":"2512.14661","title":"Focus: A Streaming Concentration Architecture for Efficient Vision-Language Models","created":"2025-12-16","cited_by_count":3,"area":"inference"},{"id":"2603.05663","title":"Keeping the Evidence Chain: Semantic Evidence Allocation for Training-Free Token Pruning in Video Temporal Grounding","created":"2026-03-05","cited_by_count":3,"area":"inference"},{"id":"2604.21335","title":"Sub-Token Routing for KV Cache Compression","created":"2026-04-23","cited_by_count":3,"area":"inference"},{"id":"2605.11605","title":"Keep What Audio Cannot Say: Context-Preserving Token Pruning for Omni-LLMs","created":"2026-05-12","cited_by_count":3,"area":"inference"},{"id":"2605.12309","title":"G$^2$TR: Generation-Guided Visual Token Reduction for Separate-Encoder Unified Multimodal Models","created":"2026-05-12","cited_by_count":3,"area":"inference"},{"id":"2606.02161","title":"InfoMerge: Information-aware Token Compression for Efficient Video Large Language Models","created":"2026-06-01","cited_by_count":3,"area":"inference"},{"id":"2608.03812","title":"OmniPack: Unified Token Compression for Efficient Omni-modal Large Language Models","created":"2026-08-04","cited_by_count":3,"area":"inference"},{"id":"2503.23959","title":"Local Information Matters: Inference Acceleration For Grounded Conversation Generation Models Through Adaptive Local-Aware Token Pruning","created":"2025-03-31","cited_by_count":2,"area":"inference"},{"id":"2603.27650","title":"V-CAST: Video Curvature-Aware Spatio-Temporal Pruning for Efficient Video Large Language Models","created":"2026-03-29","cited_by_count":2,"area":"inference"},{"id":"2607.02484","title":"Combating Textual Noise and Redundancy: Entropy-Aware Dense Visual Token Pruning","created":"2026-07-02","cited_by_count":2,"area":"inference"},{"id":"2607.03050","title":"OmniFocus: Query-Guided Modality-Balanced Token Compression for Omni-Modal Large Language Models","created":"2026-07-03","cited_by_count":2,"area":"inference"},{"id":"2604.03414","title":"KiToke: Kernel-based Interval-aware Token Compression for Video Large Language Models","created":"2026-04-03","cited_by_count":1,"area":"inference"},{"id":"2604.13432","title":"MaMe & MaRe: Matrix-Based Token Merging and Restoration for Efficient Visual Perception and Synthesis","created":"2026-04-15","cited_by_count":1,"area":"inference"},{"id":"2604.23717","title":"HeadRouter: Dynamic Head-Weight Routing for Task-Adaptive Audio Token Pruning in Large Audio Language Models","created":"2026-04-26","cited_by_count":1,"area":"inference"},{"id":"2605.11803","title":"OTT-Vid: Optimal Transport Temporal Token Compression for Video Large Language Models","created":"2026-05-12","cited_by_count":1,"area":"inference"},{"id":"2607.23193","title":"OmniScope: Modality-Decoupled Token Compression for Omnimodal Large Language Models","created":"2026-07-25","cited_by_count":1,"area":"inference"},{"id":"2607.25669","title":"OmniDelta: Skill-Driven Budget Allocation for Token Compression in OmniLLMs","created":"2026-07-28","cited_by_count":1,"area":"inference"},{"id":"2608.01665","title":"Allocation Before Ranking: Decoupled Token Compression for OmniLLMs","created":"2026-08-03","cited_by_count":1,"area":"inference"},{"id":"2511.18920","title":"EventSTU: Event-Guided Efficient Spatio-Temporal Understanding for Video Large Language Models","created":"2025-11-24","cited_by_count":0,"area":"inference"},{"id":"2604.09547","title":"Tango: Taming Visual Signals for Efficient Video Large Language Models","created":"2026-04-10","cited_by_count":0,"area":"inference"},{"id":"2604.16883","title":"SinkRouter: Sink-Aware Routing for Efficient Long-Context Decoding in Large Language and Multimodal Models","created":"2026-04-18","cited_by_count":0,"area":"inference"},{"id":"2605.16366","title":"Fre-Res: Frequency-Residual Video Token Compression for Efficient Video MLLMs","created":"2026-05-10","cited_by_count":0,"area":"inference"},{"id":"2605.30904","title":"MergeTok: Unified Continuous and Discrete Visual Tokenization via Token Merging","created":"2026-05-29","cited_by_count":0,"area":"inference"},{"id":"2607.09080","title":"GeoTrace: Geometry-Aware Trajectory Token Compression for Video Large Language Models","created":"2026-07-10","cited_by_count":0,"area":"inference"},{"id":"2607.15563","title":"Are All Tokens Necessary for Visual Place Recognition? An Empirical Study of Token Reduction for Efficient Inference","created":"2026-07-17","cited_by_count":0,"area":"inference"},{"id":"2608.01638","title":"Dynamic Resolution Routing for Efficient Egocentric Grounding","created":"2026-08-03","cited_by_count":0,"area":"inference"},{"id":"2608.08569","title":"VoxZip: Semantic-Anchored Temporal KV Cache Compression for Long-Context Audio Inference","created":"2026-08-09","cited_by_count":0,"area":"inference"},{"id":"2608.26142","title":"Position Is All You Need: A Free Lunch Token Compression Strategy for MLLM-based Referring Expression Segmentation","created":"2026-06-26","cited_by_count":0,"area":"inference"},{"id":"2608.29291","title":"Accelerating Unified Multimodal Models with Core-Expansion Routing and Unified Computation Scheduling","created":"2026-08-29","cited_by_count":0,"area":"inference"},{"id":"2609.03318","title":"Iapetus: Content-Aware Hierarchical Scheduling for Collaborative ViT Inference in LEO Satellite Networks","created":"2026-09-03","cited_by_count":0,"area":"inference"},{"id":"2609.08345","title":"CoVeR: Coverage-Based Token Pruning for Multi-View 3D Reasoning in VLMs","created":"2026-09-08","cited_by_count":0,"area":"inference"},{"id":"2609.33397","title":"CoViST: Visual Token Compression via Composable States","created":"2026-09-27","cited_by_count":0,"area":"inference"},{"id":"2609.34653","title":"OmniTide: Co-Designing Algorithms and Systems for Efficient On-Device Omni-LLM Streaming","created":"2026-09-28","cited_by_count":0,"area":"inference"},{"id":"2609.37042","title":"GleanVID: Complementary Token Selection for Efficient Video Large Language Models","created":"2026-09-29","cited_by_count":0,"area":"inference"},{"id":"2609.39704","title":"When Masking Helps or Hurts Robustness in Compressed CLIP: A Pre-Deployment Diagnostic","created":"2026-09-30","cited_by_count":0,"area":"inference"},{"id":"2610.00204","title":"Query Independent Variable Rate Visual Token Coding","created":"2026-09-20","cited_by_count":0,"area":"inference"},{"id":"2610.12959","title":"Video Codec-Guided Training-Free Token Compression for Video Large Language Models","created":"2026-10-09","cited_by_count":0,"area":"inference"},{"id":"2610.16744","title":"FoveaTree: Information-Density Trees for Query-Guided Visual Token Pruning in Video Large Language Models","created":"2026-10-12","cited_by_count":0,"area":"inference"},{"id":"2610.17355","title":"WIRE: Weight Space Implicit Representation Encoding of Frozen Video LLMs","created":"2026-10-12","cited_by_count":0,"area":"inference"},{"id":"2610.18690","title":"Just Respect the Semantics: Image-Token Compression via Pre-LLM Semantic Variable-Resolution Tokenization and In-LLM Relevance Pruning","created":"2026-10-13","cited_by_count":0,"area":"inference"}],"next_cursor":null}