Fre-Res: Frequency-Residual Video Token Compression for Efficient Video MLLMs
College of Computer Science, National University of Defense Technology（计算机科学学院，国防科技大学）；Shien-Ming Wu School of Intelligent Engineering, South China University of Technology（智能工程谢民明伍学院，华南理工大学）
Video MLLMs face a persistent tension between spatial fidelity and temporal coverage: preserving fine-grained visual details requires many spatial tokens, while capturing short-lived events requires dense temporal sampling. We propose \textbf{Fre-Res}, a budget-adaptive dual-track video-token compression framework that separates these two forms of evidence. Fre-Res preserves sparse high-fidelity spatial anchors and represents dense temporal evolution through compact residual-frequency tokens. Specifically, it applies temporal 1D-DCT to inter-frame residual trajectories in vision-latent space, where we observe strong low-frequency concentration. To align frequency-domain dynamics with native visual embeddings, Fre-Res introduces a Spatial-Guided Absorber that injects temporal residual information into spatially corresponding anchor tokens. Across fine-grained short-video and long-video reasoning benchmarks, Fre-Res achieves a favorable accuracy--efficiency trade-off, matching or approaching full-token performance while substantially reducing visual-token length. Extensive ablations further show that temporal-frequency residuals preserve causal transition cues, while spatial anchors remain essential for fine-grained object and layout reasoning.
Beyond RGB: Benchmarking and Enhancing MLLMs for Hyperspectral Image Understanding via Training-Free Reasoning Framework
Sun Yat-sen University（中山大学）；Tsinghua Shenzhen International Graduate School（清华大学深圳国际研究生院）；China Agricultural University（中国农业大学）；Southwest Jiaotong University（西南交通大学）；Southwest University（西南大学）；National Supercomputing Center in Shenzhen（国家超级计算深圳中心）
Enhancing Perception Capabilities of Multimodal LLMs with Training-Free Fusion
ZIP Lab, Zhejiang University（浙江大学 ZIP 实验室）；South China University of Technology（华南理工大学）；Pazhou Lab（琶洲实验室）；Key Laboratory of Big Data and Intelligent Robot, Ministry of Education（大数据与智能机器人教育部重点实验室）
Advancing MLLM-based UAV Image Understanding and Reasoning: A Benchmark and a Training-Free Multi-Agent System
Fudan University（复旦大学）；College of Future Information Technology（未来信息技术学院）；Shanghai Innovation Institute（上海创新研究院）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；The Chinese University of Hong Kong（香港中文大学）
本研究构建UAVQA-Bench基准，识别MLLM用于无人机图像理解的三类失效模式，提出含DSPE、CAIR、DAAS的无训练多智能体系统UAV-MAS，其32B版本在基准上准确率超Gemini 3 Pro 4.0个百分点
TAMEing Long Contexts in Personalization: Towards Training-Free and State-Aware MLLM Personalized Assistant
University of Electronic Science and Technology of China（电子科技大学）；Aiwen Technology Co., Ltd.（Aiwen科技有限公司）；Intelligent Digital Media Technology Key Laboratory of Sichuan Province（四川省智能数字媒体技术重点实验室）
School of Software, Tsinghua University（清华大学软件学院）；BNRist, Tsinghua University（清华大学BNRist）；Department of Automation, Tsinghua University（清华大学自动化系）
Multimodal Large Language Models (MLLMs) have recently demonstrated strong performance across a wide range of vision-language tasks, garnering significant attention in the computer vision. However, their efficient deployment remains a substantial challenge due to high computational costs and memory requirements. Recognizing the redundancy of information within the vision modality, recent studies have explored methods for compressing visual tokens in MLLMs to enhance efficiency in a training-free manner. Despite their effectiveness, existing methods like Fast rely on the attention between visua…
Free-MoRef: Instantly Multiplexing Context Perception Capabilities of Video-MLLMs within Single Inference
Sun Yat-sen University（中山大学）；Peng Cheng Laboratory（鹏城实验室）；OPPO AI Center（OPPO人工智能中心）；Research Institute（研究 institute）；The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））；Harbin Institute of Technology（哈尔滨工业大学）；Shenzhen Key Laboratory of Digital Living Network and Content Service（深圳数字生活网络与内容服务重点实验室）；Guangdong Key Laboratory of Big Data Analysis and Processing（广东省大数据分析与处理重点实验室）
Stabilizing Unsupervised Self-Evolution of MLLMs via Continuous Softened Retracing reSampling
Tsinghua University（清华大学）；Hefei University of Technology（合肥工业大学）；University of Arizona（亚利桑那大学）；MAIS, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所多模态人工智能系统实验室）
VFA: Empowering Multilingual MLLMs via Vision-Free Adaptation
Southern University of Science and Technology（南方科技大学）；Microsoft Research Asia（微软亚洲研究院）；Shanghai University of Finance and Economics（上海财经大学）；Peking University（北京大学）
Is Our Benchmark Enough? An Analysis of Continual Learning for MLLMs
School of Computer Science and Statistics, Trinity College Dublin（都柏林圣三一学院计算机科学与统计学院）；Department of Mathematics and Computer Science, Eindhoven University of Technology（埃因霍温理工大学数学与计算机科学系）
Balancing Efficiency and Efficacy: Training-Free Attention-Guided Switching Between Explicit and Latent Thoughts for MLLMs
Harbin Institute of Technology, Shenzhen（哈尔滨工业大学（深圳））；Tsinghua Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院）；Peng Cheng Laboratory（鹏城实验室）；Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences（中国科学院深圳先进技术研究院）
One MLLM, One Call: Efficient Zero-Shot Vision-and-Language Navigation via Spatial-Aware Waypoints
College of Electronics and Information Engineering, Shenzhen University（深圳大学电子与信息工程学院）；School of Computer Science, The University of Sydney（悉尼大学计算机科学学院）；College of Computer Science and Software Engineering, Shenzhen University（深圳大学计算机与软件学院）
