多模态大模型

Backdoor Attacks on Open Vocabulary Object Detectors via Multi-Modal Prompt Tuning

通过多模态提示调优对开放词汇目标检测器进行后门攻击

Ankita Raj, Chetan Arora

专题命中 跨模态检索 ：multi-modal(title,abstract)；分类 cs.CV

AI总结 TrAP通过多模态提示调优对开放词汇目标检测器实施后门攻击，利用轻量级提示标记植入恶意行为，提升攻击成功率并改进下游任务性能。

Comments Accepted to AAAI 2026

Enhancing Image Quality Assessment Ability of LMMs via Retrieval-Augmented Generation

通过检索增强生成提升大型多模态模型的图像质量评估能力

Kang Fu, Huiyu Duan, Zicheng Zhang, Yucheng Zhu, Jun Zhao, Xiongkuo Min, Jia Wang, Guangtao Zhai

机构 * Shanghai Jiao Tong University（上海交通大学） ； Tencent（腾讯） 纠错

专题命中 跨模态检索 ：multimodal(abstract)；分类 cs.CV、cs.AI

AI总结 IQA-RAG通过检索增强生成提升LMMs图像质量评估能力，提供高效替代微调方案

Memory in the Age of AI Agents

人工智能代理时代的记忆

Yuyang Hu, Shichun Liu, Yanwei Yue, Guibin Zhang, Boyang Liu, Fangyi Zhu, Jiahang Lin, Honglin Guo, Shihan Dou, Zhiheng Xi, Senjie Jin, Jiejun Tan, Yanbin Yin, Jiongnan Liu, Zeyu Zhang, Zhongxiang Sun, Yutao Zhu, Hao Sun, Boci Peng, Zhenrong Cheng, Xuanbo Fan, Jiaxin Guo, Xinlei Yu, Zhenhong Zhou, Zewen Hu, Jiahao Huo, Junhao Wang, Yuwei Niu, Yu Wang, Zhenfei Yin, Xiaobin Hu, Yue Liao, Qiankun Li, Kun Wang, Wangchunshu Zhou, Yixin Liu, Dawei Cheng, Qi Zhang, Tao Gui, Shirui Pan, Yan Zhang, Philip Torr, Zhicheng Dou, Ji-Rong Wen, Xuanjing Huang, Yu-Gang Jiang, Shuicheng Yan

机构 * Core Supervisors. 0.5em Affiliations: National University of Singapore, Renmin University of China, Fudan University, Peking University, Nanyang Technological University, Tongji University, University of California San Diego, Hong Kong University of Science ； Technology (Guangzhou), Griffith University, Georgia Institute of Technology, OPPO, Oxford University 纠错

专题命中 跨模态检索 ：multimodal(abstract)；分类 cs.CL、cs.AI

AI总结 本文系统梳理了人工智能代理记忆的现状与分类，提出了记忆的三种形式、功能分类及动态分析，为未来智能设计提供理论基础。