视觉大模型 / VLM - arXivDaily 专题

2507.13255 2025-09-24 cs.CL cs.AI cs.IR cs.LG cs.MM 84%

Automating Steering for Safe Multimodal Large Language Models

Lyucheng Wu, Mengru Wang, Ziwen Xu, Tri Cao, Nay Oo, Bryan Hooi, Shumin Deng

机构 * Zhejiang University（浙江大学）； Zhejiang University - Ant Group Joint Lab of Knowledge Graph（浙江大学-蚂蚁集团知识图谱联合实验室）； National University of Singapore, NUS-NCS Joint Lab（新加坡国立大学NUS-NCS联合实验室）

专题命中幻觉与鲁棒性：multimodal large language model(title,abstract)；LLaVA(abstract)；分类 cs.AI、cs.LG

Comments EMNLP 2025 Main Conference. 23 pages (8+ for main); 25 figures; 1 table

2509.19212 2025-09-24 cs.CL cs.AI 83%

Steering Multimodal Large Language Models Decoding for Context-Aware Safety

Zheyuan Liu, Zhangchen Xu, Guangyao Dou, Xiangchi Yuan, Zhaoxuan Tan, Radha Poovendran, Meng Jiang

机构 * University of Notre Dame（notre dame 大学）； University of Washington（华盛顿大学）； Johns Hopkins University（约翰霍普金斯大学）； Georgia Institute of Technology（佐治亚理工学院）

专题命中幻觉与鲁棒性：multimodal large language model(title,abstract)；MLLM(abstract)；分类 cs.AI

Comments A lightweight and model-agnostic decoding framework that dynamically adjusts token generation based on multimodal context

2509.19070 2025-09-24 cs.CV cs.CL 79%

ColorBlindnessEval: Can Vision-Language Models Pass Color Blindness Tests?

Zijian Ling, Han Zhang, Yazhuo Zhou, Jiahao Cui

机构 * Apply U United Kingdom（英国Apply大学）

专题命中幻觉与鲁棒性：vision-language model(title,abstract)；分类 cs.CV

Comments Accepted at the Open Science for Foundation Models (SCI-FM) Workshop at ICLR 2025

2505.15389 2025-09-24 cs.CL cs.CR cs.CV 79%

Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study

DongGeon Lee, Joonwon Jang, Jihae Jeong, Hwanjo Yu

机构 * Pohang University of Science and Technology (POSTECH)（釜山科学技术大学）

专题命中幻觉与鲁棒性：vision-language model(title,abstract)；分类 cs.CV

Comments Accepted to EMNLP 2025

2505.14257 2025-09-24 cs.CV 79%

Mitigating Hallucination in Large Vision-Language Models through Aligning Attention Distribution to Information Flow

Jianfei Zhao, Feng Zhang, Xin Sun, Chong Feng

机构 * School of Computer Science and Technology, Beijing Institute of Technology（计算机科学与技术学院，北京理工大学）； Zhongguancun Academy（中关村学院）； Southeast Academy of Information Technology, Beijing Institute of Technology（信息技术东南学院，北京理工大学）

专题命中幻觉与鲁棒性：vision-language model(title,abstract)；分类 cs.CV

Comments Accepted to Findings of EMNLP 2025

2509.19203 2025-09-24 cs.CV 57%

Vision-Free Retrieval: Rethinking Multimodal Search with Textual Scene Descriptions

Ioanna Ntinou, Alexandros Xenos, Yassine Ouali, Adrian Bulat, Georgios Tzimiropoulos

机构 * Queen Mary University of London（伦敦女王大学）； Samsung AI Centre（三星人工智能中心）； Technical University of Iași（伊阿苏技术大学）

专题命中幻觉与鲁棒性：vision-language model(abstract)；分类 cs.CV

Comments Accepted at EMNLP 2025

2509.18998 2025-09-24 math.AP cs.LG q-bio.CB q-bio.QM 57%

Bayesian Calibration and Model Assessment of Cell Migration Dynamics with Surrogate Model Integration

Christina Schenk, Jacobo Ayensa Jiménez, Ignacio Romero

机构 * IMDEA Materials Institute（IMDEA材料研究所）； Universidad Politécnica de Madrid（马德里理工大学）； Aragón Institute of Engineering Research (I3A)（阿伦工程技术研究所（I3A））

专题命中幻觉与鲁棒性：grounding(abstract)；分类 cs.LG

Comments 31 pages, 13 figures, 1 table

1. 幻觉与鲁棒性 7 篇

Automating Steering for Safe Multimodal Large Language Models

Steering Multimodal Large Language Models Decoding for Context-Aware Safety

ColorBlindnessEval: Can Vision-Language Models Pass Color Blindness Tests?

Are Vision-Language Models Safe in the Wild? A Meme-Based Benchmark Study

Mitigating Hallucination in Large Vision-Language Models through Aligning Attention Distribution to Information Flow

Vision-Free Retrieval: Rethinking Multimodal Search with Textual Scene Descriptions

Bayesian Calibration and Model Assessment of Cell Migration Dynamics with Surrogate Model Integration