arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of Science and Technology of China(中国科学技术大学)

2026-01-21 至 2026-01-21 共收录 19
2601.14060 2026-01-21 cs.CV

Fine-Grained Zero-Shot Composed Image Retrieval with Complementary Visual-Semantic Integration

细粒度零样本组合图像检索与互补视觉-语义整合

Yongcong Ye, Kai Zhang, Yanghai Zhang, Enhong Chen, Longfei Li, Jun Zhou

机构 * State Key Laboratory of Cognitive Intelligence, University of Science and Technology of China(认知智能国家重点实验室,中国科学技术大学) Zhejiang University(浙江大学)

AI总结 本文提出CVSI方法,通过互补视觉-语义整合提升细粒度零样本组合图像检索性能。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13995 2026-01-21 cs.CL

From Tags to Trees: Structuring Fine-Grained Knowledge for Controllable Data Selection in LLM Instruction Tuning

从标签到树:为LLM指令微调中的可控数据选择构建细粒度知识

Zihan Niu, Wenping Hu, Junmin Chen, Xiyue Wang, Tong Xu, Ruiming Tang

机构 * University of Science and Technology of China(中国科学技术大学) Klear Team, Kuaishou Technology(快手科技Klear团队)

AI总结 TAGS通过构建细粒度知识树,实现LLM指令微调中可控数据选择的高效采样与知识对齐。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13938 2026-01-21 cs.IR cs.AI

IF-GEO: Conflict-Aware Instruction Fusion for Multi-Query Generative Engine Optimization

IF-GEO:面向多查询生成引擎优化的冲突感知指令融合

Heyang Zhou, JiaJia Chen, Xiaolu Chen, Jie Bao, Zhen Chen, Yong Liao

机构 * School of Cyber Science and Technology, University of Science and Technology of China(中国科学技术大学网络科学与技术学院) Institute of Dataspace, Hefei Comprehensive National Science Center(合肥综合性国家科学中心数据研究所)

AI总结 IF-GEO通过冲突感知指令融合框架,提升多查询生成引擎在有限预算下的优化稳定性与性能。

Comments 9 pages, 3 figures. Submitted to ACL 2026. Corresponding author: Zhen Chen

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13630 2026-01-21 cs.CL

Activation-Space Anchored Access Control for Multi-Class Permission Reasoning in Large Language Models

基于激活空间锚定的多类权限推理访问控制

Zhaopeng Zhang, Pengcheng Sun, Lan Zhang, Chen Tang, Jiewei Lai, Yunhao Wang, Hui Jin

机构 * University of Science and Technology of China(中国科学技术大学) Lenovo Research(联想研究院)

AI总结 本文提出AAAC框架,通过激活空间锚点实现多类权限控制,有效降低权限违规和攻击成功率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13384 2026-01-21 cs.SE cs.CL

From Completion to Editing: Unlocking Context-Aware Code Infilling via Search-and-Replace Instruction Tuning

从补全到编辑:通过搜索和替换指令微调解锁上下文感知的代码填充

Jiajun Zhang, Zeyu Cui, Jiaxi Yang, Lei Zhang, Yuheng Jing, Zeyao Ma, Tianyi Bai, Zilei Wang, Qiang Liu, Liang Wang, Binyuan Hui, Junyang Lin

机构 * USTC(University of Science and Technology of China) Alibaba Group(阿里巴巴集团) CASIA(Chinese Academy of Sciences Institute of Automation) SIAT(Institute of Automation, Chinese Academy of Sciences)

AI总结 通过搜索和替换指令微调,SRI框架实现了上下文感知的代码填充,提升了补全性能并保持低延迟。

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.07593 2026-01-21 cs.AI

AgentAsk: Multi-Agent Systems Need to Ask

AgentAsk: 多智能体系统需要提问

Bohan Lin, Kuo Yang, Zelin Tan, Yingchuan Lai, Chen Zhang, Guibin Zhang, Xinlei Yu, Miao Yu, Xu Wang, Yudong Zhang, Yang Wang

机构 * University of Science and Technology of China(科学技术大学) Shanghai AI Laboratory(上海人工智能实验室) Xi’an Jiaotong University(西安交通大学) National University of Singapore(新加坡国立大学)

AI总结 AgentAsk通过边缘级澄清模块提升多智能体系统准确性,减少误差传播,提高任务效率与性能。

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.14833 2026-01-21 cs.CV cs.AI

Paired Image Generation with Diffusion-Guided Diffusion Models

配对图像生成与扩散引导扩散模型

Haoxuan Zhang, Wenju Cui, Yuzhu Cao, Tao Tan, Jie Liu, Yunsong Peng, Jian Zheng

机构 * University of Science and Technology of China(科学技术大学) Suzhou Institute of Biomedical Engineering and Technology(生物医学工程与技术研究所) Macao Polytechnic University(澳门 polytechnic 大学) Suzhou Municipal Hospital(苏州 municipal 医院) Guizhou Provincial People's Hospital(贵州省人民医院)

AI总结 本文提出一种无需外部条件的配对图像生成方法,通过训练额外的扩散引导器提升生成质量并缓解标注数据不足问题。

详情

展开后加载摘要…

URL PDF HTML 收藏
2502.04345 2026-01-21 cs.CL cs.AI cs.LG

Jingfang: An LLM-Based Multi-Agent System for Precise Medical Consultation and Syndrome Differentiation in Traditional Chinese Medicine

Jingfang: 一种基于大语言模型的多智能体系统,用于精准中医诊疗和辨证论治

Yehan Yang, Tianhao Ma, Ruotai Li, Xinhan Zheng, Guodong Shan

机构 * Beijing University of Posts and Telecommunications(北京邮电大学) Southeast University(东南大学) University of Chinese Academy of Sciences(中国科学院大学) University of Science and Technology of China(中国科学技术大学)

AI总结 JingFang 是一种基于大语言模型的多智能体系统,旨在通过多智能体协作机制提升中医诊疗的准确性和个性化水平,尤其在辨证论治方面表现优异。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.13007 2026-01-21 cs.SE cs.AI

ArchAgent: Scalable Legacy Software Architecture Recovery with LLMs

ArchAgent: 基于LLM的可扩展遗留软件架构恢复

Rusheng Pan, Bingcheng Mao, Tianyi Ma, Zhenhua Ling

机构 * HiThink Research(慧思研究) University of Science and Technology of China(中国科学技术大学)

AI总结 ArchAgent通过结合静态分析、自适应分段和LLM合成,实现大规模遗留软件架构的高效恢复与多视图业务对齐。

Comments to be published in ICASSP 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.12917 2026-01-21 cs.LG cs.DC

CooperLLM: Cloud-Edge-End Cooperative Federated Fine-tuning for LLMs via ZOO-based Gradient Correction

CooperLLM: 通过基于ZOO的梯度校正实现云-边-端协同的联邦微调

He Sun, Jinrui Zhou, Li Li, Mingjun Xiao

机构 * University of Science and Technology of China(中国科学技术大学) University of Macau(澳门大学)

AI总结 CooperLLM通过结合移动端ZOO和云引导梯度校正,实现云-边-端协同联邦微调,显著提升收敛速度和精度,同时保护隐私。

Comments 14 pages, 9 figures, under review

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.12428 2026-01-21 cs.RO cs.CV

ReWorld: Multi-Dimensional Reward Modeling for Embodied World Models

ReWorld:面向具身世界模型的多维奖励建模

Baorui Peng, Wenyao Zhang, Liang Xu, Zekun Qi, Jiazhao Zhang, Hongsi Liu, Wenjun Zeng, Xin Jin

机构 * Eastern Institute of Technology(东部技术研究所) Georgia Institute of Technology(佐治亚理工学院) Shanghai Jiao Tong University(上海交通大学) Tsinghua University(清华大学) University of Science and Technology of China(中国科学技术大学) Peking University(北京大学)

AI总结 ReWorld通过多维奖励建模提升具身世界模型的物理真实性和任务完成能力

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.09527 2026-01-21 cs.CV

Generative Diffusion Contrastive Network for Multi-View Clustering

多视图聚类的生成扩散对比网络

Jian Zhu, Xin Zou, Xi Wang, Lei Liu, Chang Tang, Li-Rong Dai

机构 * Zhejiang Lab(浙江实验室) Hong Kong University of Science and Technology(香港科技大学) University of Science and Technology of China(中国科学技术大学) Huazhong University of Science and Technology(华中科技大学)

AI总结 本文提出生成扩散对比网络GDCN,通过多重生成机制解决多视图聚类中的低质量数据问题,实现深度多视图聚类任务的最优性能。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.12288 2026-01-21 cs.LG cs.AI

TimeGMM: Single-Pass Probabilistic Forecasting via Adaptive Gaussian Mixture Models with Reversible Normalization

TimeGMM: 通过自适应高斯混合模型与可逆归一化实现单次传递概率预测

Lei Liu, Tengyuan Liu, Hongwei Zhao, Jiahui Huang, Ruibo Guo, Bin Li

机构 * University of Science and Technology of China(中国科学技术大学)

AI总结 TimeGMM通过自适应高斯混合模型与可逆归一化实现单次传递概率预测,有效提升预测精度和分布匹配度。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.12252 2026-01-21 cs.HC cs.CV

Breaking Coordinate Overfitting: Geometry-Aware WiFi Sensing for Cross-Layout 3D Pose Estimation

打破坐标过拟合:面向跨布局3D姿态估计的几何感知WiFi传感

Songming Jia, Yan Lu, Bin Liu, Xiang Zhang, Peng Zhao, Xinmeng Tang, Yelin Wei, Jinyang Huang, Huan Yan, Zhi Liu

机构 * University of Science and Technology of China(科学技术大学) Shanghai Artificial Intelligence Lab(上海人工智能实验室) Tianjin University(天津大学) Hefei University of Technology(合肥工业大学) Guizhou Normal University(贵州师范大学) The University of Electro-Communications(电子通信大学)

AI总结 PerceptAlign通过几何条件化学习实现布局无关的WiFi姿态估计,显著提升跨域泛化能力。

Comments Accpeted by AMC Mobicom 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.12091 2026-01-21 cs.LG

Mitigating Cultural Bias in LLMs via Multi-Agent Cultural Debate

通过多智能体文化辩论缓解LLM中的文化偏见

Qian Tan, Lei Jiang, Yuting Zeng, Shuoyang Ding, Xiaohua Xu

机构 * University of Science and Technology of China(中国科学技术大学) NVIDIA(英伟达)

AI总结 通过多智能体文化辩论框架,有效缓解LLM中的文化偏见,提升跨文化公平性。

Comments 13 pages

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.10688 2026-01-21 cs.IR cs.AI

Rethinking Popularity Bias in Collaborative Filtering via Analytical Vector Decomposition

重新思考协同过滤中的流行度偏差:通过分析向量分解

Lingfeng Liu, Yixin Song, Dazhong Shen, Bing Yin, Hao Li, Yanyong Zhang, Chao Wang

机构 * School of Artificial Intelligence and Data Science, University of Science and Technology of China(人工智能与数据科学学院,科学技术大学) College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics(计算机科学与技术学院,南京航空航天大学) State Key Laboratory of Cognitive Intelligence(认知智能国家重点实验室) iFLYTEK Research, iFLYTEK(iFLYTEK研究)

AI总结 本文提出DDC框架,通过不对称方向更新修正协同过滤中流行度偏差的几何问题,提升推荐质量和公平性。

Comments Accepted by SIGKDD 2026(First Cycle)

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.18530 2026-01-21 cs.SD eess.AS

A Stage-Wise Learning Strategy with Fixed Anchors for Robust Speaker Verification

具有固定锚点的分阶段学习策略用于鲁棒语音验证

Bin Gu, Lipeng Dai, Huipeng Du, Haitao Zhao, Jibo Wei

机构 * National University of Defense Technology, Changsha, China(国防科技大学,长沙,中国) University of Science and Technology of China, Hefei, China(中国科学技术大学,合肥,中国)

AI总结 本文提出一种基于固定锚点的分阶段学习策略,用于在噪声条件下提高语音验证的鲁棒性和判别性。

Comments submitted to ICASSP 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.06467 2026-01-21 cs.CV

Does DINOv3 Set a New Medical Vision Standard? Benchmarking 2D and 3D Classification, Segmentation, and Registration

DINOv3 是否设定了医学视觉的新标准?对2D和3D分类、分割与配准的基准测试

Che Liu, Yinda Chen, Haoyuan Shi, Jinpeng Lu, Bailiang Jian, Jiazhen Pan, Linghan Cai, Jiayi Wang, Jieming Yu, Ziqi Gao, Xiaoran Zhang, Long Bai, Yundi Zhang, Jun Li, Cosmin I. Bercea, Cheng Ouyang, Chen Chen, Zhiwei Xiong, Benedikt Wiestler, Christian Wachinger, James S. Duncan, Daniel Rueckert, Wenjia Bai, Rossella Arcucci

机构 * Imperial College London(伦敦帝国理工学院) University of Science and Technology of China(中国科学技术大学) Dresden University of Technology(德累斯顿技术大学) University of Erlangen-Nuremberg(埃尔兰根-纽伦堡大学) University of Oxford(牛津大学) University of Sheffield(谢菲尔德大学) Technical University of Munich (TUM)(慕尼黑技术大学) Munich Center for Machine Learning(慕尼黑机器学习中心) The Hong Kong University of Science and Technology(香港科学与技术大学) The Chinese University of Hong Kong(香港中文大学) Yale University(耶鲁大学)

AI总结 DINOv3在医学视觉任务中表现出色,但其在深度领域专门化任务中存在性能退化问题。

Comments Technical Report

详情

展开后加载摘要…

URL PDF HTML 收藏
2412.03938 2026-01-21 cs.LG cs.CR

JANUS: A Difference-Oriented Analyzer For Financial Centralization Risks in Smart Contracts

JANUS:一种面向智能合约金融集中风险的差分分析器

Wansen Wang, Pu Zhang, Renjie Ji, Wenchao Huang, Zhaoyi Meng, Yan Xiong

机构 * School of Computer Science and Technology, Anhui University(安徽大学计算机科学与技术学院) School of Computer Science and Technology, University of Science and Technology of China(中国科学技术大学计算机科学与技术学院)

AI总结 JANUS通过分析智能合约中特权账户与普通账户状态差异,检测金融集中风险,提高了检测准确性并发现未知模式的风险。

详情

展开后加载摘要…

URL PDF HTML 收藏