AI 大模型

视觉大模型 / VLM

视觉语言模型、视觉推理、视觉问答、图文理解和视觉 grounding。

2025-10-24 至 2025-10-24 共收录 5 篇信号源：cs.CV, cs.AI, cs.LG

1. VLM训练与架构 5 篇

2510.14349 2025-10-24 cs.CV 83%

Vision-Centric Activation and Coordination for Multimodal Large Language Models

Yunnan Wang, Fan Lu, Kecheng Zheng, Ziyuan Huang, Ziqiang Li, Wenjun Zeng, Xin Jin

机构 * MoE Key Lab of Artificial Intelligence, Shanghai Jiao Tong University（人工智能MoE实验室，上海交通大学）； Ant Group（蚂蚁集团）； Ningbo Institute of Digital Twin, Eastern Institute of Technology, Ningbo（宁波数字孪生研究所，东部技术研究所，宁波）

专题命中 VLM训练与架构：multimodal large language model(title,abstract)；MLLM(abstract)；分类 cs.CV

2509.15235 2025-10-24 cs.CV cs.CL 83%

ViSpec: Accelerating Vision-Language Models with Vision-Aware Speculative Decoding

Jialiang Kang, Han Shu, Wenshuo Li, Yingjie Zhai, Xinghao Chen

机构 * Peking University（北京大学）； Huawei Noah’s Ark Lab（华为诺亚实验室）

专题命中 VLM训练与架构：vision-language model(title,abstract)；VLM(abstract)；分类 cs.CV

Comments NeurIPS 2025

2505.22038 2025-10-24 cs.CV cs.AI 81%

Balanced Token Pruning: Accelerating Vision Language Models Beyond Local Optimization

Kaiyuan Li, Xiaoyue Chen, Chen Gao, Yong Li, Xinlei Chen

机构 * Tsinghua Shenzhen International Graduate School（清华大学深圳国际研究生院）； BNRist, Tsinghua University（清华大学北京研究院）

专题命中 VLM训练与架构：vision language model(title)；vision-language model(abstract)；分类 cs.CV、cs.AI

Comments Accepted by Neurips 2025

2510.20639 2025-10-24 cs.CV 79%

Better Tokens for Better 3D: Advancing Vision-Language Modeling in 3D Medical Imaging

Ibrahim Ethem Hamamci, Sezgin Er, Suprosanna Shit, Hadrien Reynaud, Dong Yang, Pengfei Guo, Marc Edgar, Daguang Xu, Bernhard Kainz, Bjoern Menze

机构 * University of Zurich（苏黎世大学）； NVIDIA ； Imperial College London（伦敦帝国理工学院）； FAU Erlangen-Nürnberg（埃尔朗根-纽伦堡大学）

专题命中 VLM训练与架构：vision-language model(title,abstract)；分类 cs.CV

Comments NeurIPS 2025

2505.19201 2025-10-24 cs.CL 67%

DREAM: Drafting with Refined Target Features and Entropy-Adaptive Cross-Attention Fusion for Multimodal Speculative Decoding

Yunhai Hu, Tianhua Xia, Zining Liu, Rahul Raman, Xingyu Liu, Bo Bao, Eric Sather, Vithursan Thangarasa, Sai Qian Zhang

机构 * Courant Institute of Mathematical Sciences, New York University（纽约大学数学科学学院）； Tandon School of Engineering, New York University（纽约大学工程学院）； Cerebras Systems Inc.（Cerebras Systems公司）； University of Pennsylvania（宾夕法尼亚大学）

专题命中 VLM训练与架构：vision-language model(abstract)；LLaVA(abstract)