{"total":95,"rows":[{"id":"2407.02490","title":"MInference 1.0: Accelerating Pre-filling for Long-Context LLMs via Dynamic Sparse Attention","created":"2024-07-02","cited_by_count":383,"area":null},{"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":"2412.19442","title":"A Survey on Large Language Model Acceleration based on KV Cache Management","created":"2024-12-27","cited_by_count":131,"area":"lr"},{"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":"2506.01713","title":"SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning","created":"2025-06-02","cited_by_count":57,"area":"lr"},{"id":"2503.12559","title":"AdaReTaKe: Adaptive Redundancy Reduction to Perceive Longer for Video-language Understanding","created":"2025-03-16","cited_by_count":48,"area":"mllm"},{"id":"2410.08584","title":"ZipVL: Efficient Large Vision-Language Models with Dynamic Token Sparsification","created":"2024-10-11","cited_by_count":44,"area":"mllm"},{"id":"2410.11842","title":"MoH: Multi-Head Attention as Mixture-of-Head Attention","created":"2024-10-15","cited_by_count":44,"area":"moe"},{"id":"2510.18269","title":"StreamingTOM: Streaming Token Compression for Efficient Video Understanding","created":"2025-10-21","cited_by_count":43,"area":"mllm"},{"id":"2412.13180","title":"Feather the Throttle: Revisiting Visual Token Pruning for Vision-Language Model Acceleration","created":"2024-12-17","cited_by_count":41,"area":"mllm"},{"id":"2502.04420","title":"KVTuner: Sensitivity-Aware Layer-Wise Mixed-Precision KV Cache Quantization for Efficient and Nearly Lossless LLM Inference","created":"2025-02-06","cited_by_count":40,"area":"inference"},{"id":"2409.17422","title":"Discovering the Gems in Early Layers: Accelerating Long-Context LLMs with 1000x Input Token Reduction","created":"2024-09-25","cited_by_count":39,"area":"inference"},{"id":"2410.16267","title":"xGen-MM-Vid (BLIP-3-Video): You Only Need 32 Tokens to Represent a Video Even in VLMs","created":"2024-10-21","cited_by_count":34,"area":"mllm"},{"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":"2502.17599","title":"MEDA: Dynamic KV Cache Allocation for Efficient Multimodal Long-Context Inference","created":"2025-02-24","cited_by_count":31,"area":"mllm"},{"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":"2505.19147","title":"Shifting AI Efficiency From Model-Centric to Data-Centric Compression","created":"2025-05-25","cited_by_count":29,"area":"mllm"},{"id":"2506.03642","title":"Spatial Understanding from Videos: Structured Prompts Meet Simulation Data","created":"2025-06-04","cited_by_count":28,"area":null},{"id":"2411.03312","title":"Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters","created":"2024-11-05","cited_by_count":25,"area":"mllm"},{"id":"2505.19602","title":"Memory-Efficient Visual Autoregressive Modeling with Scale-Aware KV Cache Compression","created":"2025-05-26","cited_by_count":22,"area":"inference"},{"id":"2412.05185","title":"LinVT: Empower Your Image-level Large Language Model to Understand Videos","created":"2024-12-06","cited_by_count":19,"area":null},{"id":"2504.09261","title":"Head-Aware KV Cache Compression for Efficient Visual Autoregressive Modeling","created":"2025-04-12","cited_by_count":15,"area":"inference"},{"id":"2502.02789","title":"Speculative Prefill: Turbocharging TTFT with Lightweight and Training-Free Token Importance Estimation","created":"2025-02-05","cited_by_count":14,"area":null},{"id":"2510.20707","title":"Mixing Importance with Diversity: Joint Optimization for KV Cache Compression in Large Vision-Language Models","created":"2025-10-23","cited_by_count":14,"area":"mllm"},{"id":"2412.03409","title":"PrefixKV: Adaptive Prefix KV Cache is What Vision Instruction-Following Models Need for Efficient Generation","created":"2024-12-04","cited_by_count":13,"area":"lr"},{"id":"2503.16257","title":"Plug-and-Play 1.x-Bit KV Cache Quantization for Video Large Language Models","created":"2025-03-20","cited_by_count":13,"area":"inference"},{"id":"2412.04652","title":"Cross-Self KV Cache Pruning for Efficient Vision-Language Inference","created":"2024-12-05","cited_by_count":12,"area":"inference"},{"id":"2506.15724","title":"MadaKV: Adaptive Modality-Perception KV Cache Eviction for Efficient Multimodal Long-Context Inference","created":"2025-06-06","cited_by_count":12,"area":"mllm"},{"id":"2508.06297","title":"KV Cache Compression for Inference Efficiency in LLMs: A Review","created":"2025-08-08","cited_by_count":12,"area":"inference"},{"id":"2510.00536","title":"GUI-KV: Efficient GUI Agents via KV Cache with Spatio-Temporal Awareness","created":"2025-10-01","cited_by_count":11,"area":"mllm"},{"id":"2510.07318","title":"Artificial Hippocampus Networks for Efficient Long-Context Modeling","created":"2025-10-08","cited_by_count":11,"area":null},{"id":"2510.16598","title":"VisionSelector: End-to-End Learnable Visual Token Compression for Efficient Multimodal LLMs","created":"2025-10-18","cited_by_count":11,"area":"mllm"},{"id":"2504.00557","title":"Efficient LLaMA-3.2-Vision by Trimming Cross-attended Visual Features","created":"2025-04-01","cited_by_count":10,"area":"mllm"},{"id":"2512.14946","title":"EVICPRESS: Joint KV-Cache Compression and Eviction for Efficient LLM Serving","created":"2025-12-16","cited_by_count":10,"area":"inference"},{"id":"2501.15021","title":"AKVQ-VL: Attention-Aware KV Cache Adaptive 2-Bit Quantization for Vision-Language Models","created":"2025-01-25","cited_by_count":9,"area":"mllm"},{"id":"2604.05546","title":"Efficient Inference for Large Vision-Language Models: Bottlenecks, Techniques, and Prospects","created":"2026-04-07","cited_by_count":9,"area":"lr"},{"id":"2409.09808","title":"Famba-V: Fast Vision Mamba with Cross-Layer Token Fusion","created":"2024-09-15","cited_by_count":8,"area":"ssm"},{"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":"2601.15655","title":"Event-VStream: Event-Driven Real-Time Understanding for Long Video Streams","created":"2026-01-22","cited_by_count":6,"area":"mllm"},{"id":"2410.03007","title":"FastAdaSP: Multitask-Adapted Efficient Inference for Large Speech Language Model","created":"2024-10-03","cited_by_count":5,"area":"inference"},{"id":"2506.02850","title":"METok: Multi-Stage Event-based Token Compression for Efficient Long Video Understanding","created":"2025-06-03","cited_by_count":5,"area":"inference"},{"id":"2507.05557","title":"Enhancing Test-Time Scaling of Large Language Models with Hierarchical Retrieval-Augmented MCTS","created":"2025-07-08","cited_by_count":5,"area":"lr"},{"id":"2601.12904","title":"From Prefix Cache to Fusion RAG Cache: Accelerating LLM Inference in Retrieval-Augmented Generation","created":"2026-01-19","cited_by_count":5,"area":"rag"},{"id":"2605.19660","title":"OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond","created":"2026-05-19","cited_by_count":5,"area":"inference"},{"id":"2412.13437","title":"Deploying Foundation Model Powered Agent Services: A Survey","created":"2024-12-18","cited_by_count":4,"area":"agents"},{"id":"2511.09833","title":"ACT as Human: Multimodal Large Language Model Data Annotation with Critical Thinking","created":"2025-11-13","cited_by_count":4,"area":"lr"},{"id":"2604.18260","title":"Geometry-Guided 3D Visual Token Pruning for Video-Language Models","created":"2026-04-20","cited_by_count":4,"area":"mllm"},{"id":"2607.08032","title":"What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents","created":"2026-07-09","cited_by_count":4,"area":"agents"},{"id":"2501.01792","title":"Efficient LLM Inference with Activation Checkpointing and Hybrid Caching","created":"2025-01-03","cited_by_count":3,"area":"inference"},{"id":"2502.14882","title":"CalibQuant: 1-Bit KV Cache Quantization for Multimodal LLMs","created":"2025-02-15","cited_by_count":3,"area":"mllm"},{"id":"2503.08019","title":"Multi-Cue Adaptive Visual Token Pruning for Large Vision-Language Models","created":"2025-03-11","cited_by_count":3,"area":"mllm"},{"id":"2503.08461","title":"FastCache: Optimizing Multimodal LLM Serving through Lightweight KV-Cache Compression Framework","created":"2025-03-11","cited_by_count":3,"area":"inference"},{"id":"2506.00479","title":"EffiVLM-BENCH: A Comprehensive Benchmark for Evaluating Training-Free Acceleration in Large Vision-Language Models","created":"2025-05-31","cited_by_count":3,"area":"mllm"},{"id":"2508.17243","title":"CoViPAL: Layer-wise Contextualized Visual Token Pruning for Large Vision-Language Models","created":"2025-08-24","cited_by_count":3,"area":"mllm"},{"id":"2511.16786","title":"Revisiting Multimodal KV Cache Compression: A Frequency-Domain-Guided Outlier-KV-Aware Approach","created":"2025-11-20","cited_by_count":3,"area":"mllm"},{"id":"2604.05887","title":"HybridKV: Hybrid KV Cache Compression for Efficient Multimodal Large Language Model Inference","created":"2026-04-07","cited_by_count":3,"area":"lr"},{"id":"2604.25642","title":"Prefill-Time Intervention for Mitigating Hallucination in Large Vision-Language Models","created":"2026-04-28","cited_by_count":3,"area":"mllm"},{"id":"2605.04075","title":"RetentiveKV: State-Space Memory for Uncertainty-Aware Multimodal KV Cache Eviction","created":"2026-04-14","cited_by_count":3,"area":"mllm"},{"id":"2605.12260","title":"PRISM: Pareto-Efficient Retrieval over Intent-Aware Structured Memory for Long-Horizon Agents","created":"2026-05-12","cited_by_count":3,"area":"agents"},{"id":"2605.17170","title":"TriAxialKV: Toward Extreme Low-Precision KV-Cache Quantization for Agentic Inference Tasks","created":"2026-05-16","cited_by_count":3,"area":"agents"},{"id":"2606.12412","title":"Reroute, Don't Remove: Recoverable Visual Token Routing for Vision-Language Models","created":"2026-06-10","cited_by_count":3,"area":"mllm"},{"id":"2606.23581","title":"Kamera: Unified Position-Invariant Multimodal KV Cache for Training-Free Reuse","created":"2026-06-22","cited_by_count":3,"area":"inference"},{"id":"2503.11205","title":"LLaVA-MLB: Mitigating and Leveraging Attention Bias for Training-Free Video LLMs","created":"2025-03-14","cited_by_count":2,"area":null},{"id":"2505.15816","title":"Streamline Without Sacrifice -- Squeeze out Computation Redundancy in LMM","created":"2025-05-21","cited_by_count":2,"area":"mllm"},{"id":"2505.22396","title":"Zooming from Context to Cue: Hierarchical Preference Optimization for Multi-Image MLLMs","created":"2025-05-28","cited_by_count":2,"area":"mllm"},{"id":"2506.16691","title":"LaVi: Efficient Large Vision-Language Models via Internal Feature Modulation","created":"2025-06-20","cited_by_count":2,"area":"mllm"},{"id":"2512.04857","title":"Autoregressive Image Generation Needs Only a Few Lines of Cached Tokens","created":"2025-12-04","cited_by_count":2,"area":"inference"},{"id":"2606.31982","title":"ERA: Entropy-Guided Visual Token Pruning with Rectified Attention for Efficient MLLMs","created":"2026-06-30","cited_by_count":2,"area":"mllm"},{"id":"2607.23265","title":"WaveZip: Wavelet-Driven Space-Time Decoupling for Video Token Condensation","created":"2026-07-25","cited_by_count":2,"area":"mllm"},{"id":"2504.21403","title":"Static or Dynamic: Towards Query-Adaptive Token Selection for Video Question Answering","created":"2025-04-30","cited_by_count":1,"area":null},{"id":"2601.20357","title":"TABED: Test-Time Adaptive Ensemble Drafting for Robust Speculative Decoding in LVLMs","created":"2026-01-28","cited_by_count":1,"area":"inference"},{"id":"2603.23483","title":"SpecEyes: Accelerating Agentic Multimodal LLMs via Speculative Perception and Planning","created":"2026-03-24","cited_by_count":1,"area":"lr"},{"id":"2605.06611","title":"The Structural Origin of Attention Sink: Variance Discrepancy, Super Neurons, and Dimension Disparity","created":"2026-05-07","cited_by_count":1,"area":null},{"id":"2605.31457","title":"VisionPulse: Dynamic Visual Sparsity for Efficient Multimodal Reasoning","created":"2026-05-29","cited_by_count":1,"area":"lr"},{"id":"2605.31598","title":"Linear Scaling Video VLMs for Long Video Understanding","created":"2026-05-29","cited_by_count":1,"area":"mllm"},{"id":"2606.01790","title":"STaR-KV: Spatio-Temporal Adaptive Re-weighting for KV Cache Compression in GUI Vision-Language Models","created":"2026-06-01","cited_by_count":1,"area":"mllm"},{"id":"2606.08302","title":"HACK++: Towards More Effective Head-Aware Key-Value Compression for Efficient Visual Autoregressive Modeling","created":"2026-06-06","cited_by_count":1,"area":null},{"id":"2606.14782","title":"Last But Not Least: Boundary Attention CalibratiON for Multimodal KV Cache Compression","created":"2026-06-10","cited_by_count":1,"area":"lr"},{"id":"2607.25467","title":"Seen, Said, or Forgotten? A Causal Audit of Visual KV Memory Across Dialog Turns","created":"2026-07-28","cited_by_count":1,"area":null},{"id":"2507.09071","title":"BlindSight: Harnessing Sparsity for Efficient Vision-Language Models","created":"2025-07-11","cited_by_count":0,"area":"mllm"},{"id":"2603.23914","title":"Attention-aware Inference Optimizations for Large Vision-Language Models with Memory-efficient Decoding","created":"2026-03-25","cited_by_count":0,"area":"mllm"},{"id":"2604.11627","title":"POINTS-Long: Adaptive Dual-Mode Visual Reasoning in MLLMs","created":"2026-04-13","cited_by_count":0,"area":"lr"},{"id":"2604.16734","title":"Reducing Peak Memory Usage for Modern Multimodal Large Language Model Pipelines","created":"2026-04-17","cited_by_count":0,"area":"mllm"},{"id":"2605.08158","title":"HY-Himmel Technical Report: Hierarchical Interleaved Multi-stream Motion Encoding for Long Video Understanding","created":"2026-05-04","cited_by_count":0,"area":"mllm"},{"id":"2605.16439","title":"KVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy","created":"2026-05-14","cited_by_count":0,"area":"lr"},{"id":"2605.19218","title":"Rotation-Aligned Key Channel Pruning for Efficient Vision-Language Model Inference","created":"2026-05-19","cited_by_count":0,"area":"mllm"},{"id":"2607.16213","title":"SelKV: Selective KV Cache Merging with Per-Token Merge-or-Drop and Attention Compensation","created":"2026-05-19","cited_by_count":0,"area":"inference"},{"id":"2607.22586","title":"MM-ShiftKV: Decode-Aware Prefill-Stage KV Selection for Multimodal Large Language Models","created":"2026-06-09","cited_by_count":0,"area":"mllm"},{"id":"2609.05821","title":"CONDUIT: A Unified Residual-Stream Restoration Framework for KV Cache Reuse in Vision-Language Models","created":"2026-09-05","cited_by_count":0,"area":"mllm"},{"id":"2609.06663","title":"ECOKV: Geometry-Aware KV Cache Eviction via Complementary Diversity Metrics","created":"2026-09-06","cited_by_count":0,"area":"mllm"},{"id":"2609.10355","title":"Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs","created":"2026-09-09","cited_by_count":0,"area":null},{"id":"2609.33334","title":"When to Evict, Not What to Keep: Draft-Guided Eviction for Training-Free KV-Cache Compression","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":"2609.38781","title":"ChartDensity-Bench: Benchmarking MLLMs for Numerical Data Reconstruction under Visual Density","created":"2026-09-30","cited_by_count":0,"area":"mllm"},{"id":"2610.17530","title":"TwinPrune: Density-Aware Two-Phase Token Pruning for Vision-Language Models","created":"2026-10-13","cited_by_count":0,"area":"mllm"}],"next_cursor":null}