token compression

Unified Spatiotemporal Token Compression for Video-LLMs at Ultra-Low Retention CVPR 2026 URaG: Unified Retrieval and Generation in Multimodal LLMs for Efficient Long Document Understanding AAAI 2026 Variation-aware Vision Token Dropping for Faster Large Vision-Language Models CVPR 2026 Vista: Scene-Aware Optimization for Streaming Video Question Answering Under Post-Hoc Queries AAAI 2026 500xCompressor: Generalized Prompt Compression for Large Language Models ACL 2025 AIM: Let Any Multimodal Large Language Models Embrace Efficient In-Context Learning AAAI 2025 AVAM: a Universal Training-free Adaptive Visual Anchoring Embedded into Multimodal Large Language Model for Multi-image Question Answering ICCV 2025 Cramming 1568 Tokens into a Single Vector and Back Again: Exploring the Limits of Embedding Space Capacity ACL 2025 CROP: Contextual Region-Oriented Visual Token Pruning EMNLP 2025 DyCoke: Dynamic Compression of Tokens for Fast Video Large Language Models CVPR 2025 Dynamic-VLM: Simple Dynamic Visual Token Compression for VideoLLM ICCV 2025 EffiVLM-BENCH: A Comprehensive Benchmark for Evaluating Training-Free Acceleration in Large Vision-Language Models ACL 2025 Enhancing Cross-Lingual Transfer through Reversible Transliteration: A Huffman-Based Approach for Low-Resource Languages ACL 2025 FOLDER: Accelerating Multi-Modal Large Language Models with Enhanced Performance ICCV 2025 Free-MoRef: Instantly Multiplexing Context Perception Capabilities of Video-MLLMs within Single Inference ICCV 2025 From Long to Lean: Performance-aware and Adaptive Chain-of-Thought Compression via Multi-round Refinement EMNLP 2025 Hybrid-Level Instruction Injection for Video Token Compression in Multi-modal Large Language Models CVPR 2025 Inference Optimal VLMs Need Fewer Visual Tokens and More Parameters ICLR 2025 Layer- and Timestep-Adaptive Differentiable Token Compression Ratios for Efficient Diffusion Transformers CVPR 2025 Libra-Merging: Importance-redundancy and Pruning-merging Trade-off for Acceleration Plug-in in Large Vision-Language Model CVPR 2025 LLaVA-Mini: Efficient Image and Video Large Multimodal Models with One Vision Token ICLR 2025 METok: Multi-Stage Event-based Token Compression for Efficient Long Video Understanding EMNLP 2025 Mini-Monkey: Alleviating the Semantic Sawtooth Effect for Lightweight MLLMs via Complementary Image Pyramid ICLR 2025 Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention ACL 2025 NVILA: Efficient Frontier Visual Language Models CVPR 2025