检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Pan, Wenbo Liu, Zhichao Chen, Qiguang Zhou, Xiangyang Yu, Haining Jia, Xiaohua Department of Computer Science City University of Hong Kong Hong Kong School of Cyberspace Science Harbin Institute of Technology China Research Center for Social Computing and Information Retrieval Harbin Institute of Technology China Microsoft United States

Large Language Models’ safety-aligned behaviors, such as refusing harmful queries, can be represented by linear directions in activation space. Previous research modeled safety behavior with a single direction, limiting mechanistic understanding to an isolated safety feature. In this work, we discover that safety-aligned behavior is jointly controlled by multi-dimensional directions. Namely, we study the vector space of representation shifts during safety fine-tuning on Llama 3 8B for refusing jailbreaks. By studying orthogonal directions in the space, we first find that a dominant direction governs the model’s refusal behavior, while multiple smaller directions represent distinct and interpretable features like hypothetical narrative and role-playing. We then measure how different directions promote or suppress the dominant direction, showing the important role of secondary directions in shaping the model’s refusal representation. Finally, we demonstrate that removing certain trigger tokens in harmful queries can mitigate these directions to bypass the learned safety capability, providing new insights on understanding safety alignment vulnerability from a multi-dimensional perspective. Code and artifacts are available at https://***/BMPixel/safety-residual-space. © 2025, CC BY.

关键词： Safety engineering

来源：评论

学校读者我要写书评

暂无评论

Multi-Granularity Distribution Alignment for Cross-Domain Crowd Counting

引用

IEEE Transactions on Image Processing 2025年 34卷 3648-3662页

作者： Zhong, Xian Qiu, Lingyue Zhu, Huilin Yuan, Jingling He, Shengfeng Wang, Zheng Wuhan University of Technology Sanya Science and Education Innovation Park Sanya 572025 China Nanyang Technological University Rapid-Rich Object Search Lab School of Electrical and Electronic Engineering 639798 Singapore Wuhan University of Technology Hubei Key Laboratory of Transportation Internet of Things School of Computer Science and Artificial Intelligence Wuhan 430070 China Wuhan University of Technology School of Computer Science and Artificial Intelligence Wuhan 430070 China Singapore Management University School of Computing and Information Systems 188065 Singapore Wuhan University National Engineering Research Center for Multimedia Software School of Computer Science Wuhan 430072 China

Unsupervised domain adaptation enables the transfer of knowledge from a labeled source domain to an unlabeled target domain, and its application in crowd counting is gaining momentum. Current methods typically align distributions across domains to address inter-domain disparities at a global level. However, these methods often struggle with significant intra-domain gaps caused by domain-agnostic factors such as density, surveillance angles, and scale, leading to inaccurate alignment and unnecessary computational burdens, especially in large-scale training scenarios. To address these challenges, we propose the Multi-Granularity Optimal Transport (MGOT) distribution alignment framework, which aligns domain-agnostic factors across domains at different granularities. The motivation behind multi-granularity is to capture fine-grained domain-agnostic variations within domains. Our method proceeds in three phases: first, clustering coarse-grained features based on intra-domain similarity;second, aligning the granular clusters using an optimal transport framework and constructing a mapping from cluster centers to finer patch levels between domains;and third, re-weighting the aligned distribution for model refinement in domain adaptation. Extensive experiments across twelve cross-domain benchmarks show that our method outperforms existing state-of-the-art methods in adaptive crowd counting. © 1992-2012 IEEE.

关键词： crowd counting distribution alignment multi-granularity optimal transport unsupervised domain adaptation

来源：评论

学校读者我要写书评

暂无评论

Data-Free Network Debiasing for Long-Tailed Visual Recognition

Data-Free Network Debiasing for Long-Tailed Visual Recogniti...

引用

2022 IEEE International Conference on Multimedia and Expo, ICME 2022

作者： Cai, Jinmian Wang, Zheng Fu, Huazhu Chen, Jingjing Jiang, Yu-Gang School of Computer Science Fudan University Shanghai Key Lab of Intelligent Information Processing China Shanghai Collaborative Innovation Center on Intelligent Visual Computing China Biren Technology IHPC A-STAR

ISBN: (数字)9781665485630

ISBN: (纸本)9781665485630

Real-world data is often unbalanced and exhibits long-tailed distribution over classes. Vanilla classification models trained on imbalanced datasets inherently exhibit bias towards dominant classes. Existing debiasing methods mostly balance the data or the loss during training. Nevertheless, these data-acquiring methods are not suitable for situations where training data are unavailable. In this paper, we appeal to solutions without access to training data and propose a datafree debiasing (Free-D) method that serves as a plug-and-play module for any standard classification model. Specifically, our method adjusts both the feature representation via feature representation shifting and the classifier weight via class prior compensation in a data-free manner. We evaluate and compare our methods on four long-tailed visual recognition datasets, i.e., long-tailed CIFAR-10/-100, ImageNet-LT, and Places-LT. Extensive experiments demonstrate that the proposed data-free method achieves comparable results of other data-acquired methods. © 2022 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

StreamFP: Learnable Fingerprint-guided Data Selection for Efficient Stream Learning

arXiv

引用

arXiv 2024年

作者： Shi, Tongjun Zhang, Shuhao Chen, Binbin He, Bingsheng National Engineering Research Center for Big DataTechnology and System Services Computing Technology and System Lab Cluster Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Singapore University of Technology and Design Singapore National University of Singapore Singapore

Stream Learning (SL) requires models that can quickly adapt to continuously evolving data, posing significant challenges in both computational efficiency and learning accuracy. Effective data selection is critical in SL to ensure a balance between information retention and training efficiency. Traditional rule-based data selection methods struggle to accommodate the dynamic nature of streaming data, highlighting the necessity for innovative solutions that effectively address these challenges. Recent approaches to handling changing data distributions face challenges that limit their effectiveness in fast-paced environments. In response, we propose StreamFP, a novel approach that uniquely employs dynamic, learnable parameters called fingerprints to enhance data selection efficiency and adaptability in stream learning. StreamFP optimizes coreset selection through its unique fingerprint-guided mechanism for efficient training while ensuring robust buffer updates that adaptively respond to data dynamics, setting it apart from existing methods in stream learning. Experimental results demonstrate that StreamFP outperforms state-of-the-art methods by achieving accuracy improvements of 15.99%, 29.65%, and 51.24% compared to baseline models across varying data arrival rates, alongside a training throughput increase of 4.6x. © 2024, CC BY.

关键词： Data accuracy

来源：评论

学校读者我要写书评

暂无评论

CADEC: A Combinatorial Auction for Dynamic Distributed DNN Inference Scheduling in Edge-Cloud Networks

引用

IEEE Transactions on Mobile computing 2025年

作者： Xu, Xiaolong Hu, Yuhao Cui, Guangming Qi, Lianyong Dou, Wanchun Nanjing University of Information Science and Technology School of Software Nanjing210044 China Nanjing University of Information Science and Technology Jiangsu Province Engineering Research Center of Advanced Computing and Intelligent Services School of Software China Nanjing University State Key Laboratory for Novel Software Technology China College of Computer science and technology Qingdao China Georgia State University Department of Computer Science United States

Deep Neural Network (DNN) Inference, as a key enabler of intelligent applications, is often computation-intensive and latency-sensitive. Combining the advantages of cloud computing (abundant computing resources) and edge computing (fast transmission), edge-cloud collaborative DNN inference is a powerful solution to these problems. However, in edge-cloud networks with heterogeneous resources, how to obtain reasonable decisions on server selection, model partition and resource allocation for efficient distributed DNN inference is a hard challenge. Furthermore, it is non-trivial to design suitable resource prices to maximize the social welfare. These challenges even escalate in dynamic edge-cloud networks where decisions should be generated as soon as each user arrives without future information. Therefore, we design a combinatorial auction for dynamic distributed DNN inference scheduling, named CADEC. CADEC first constructs a bid set for each user based on convex optimization theory for optimal solution searching. Next, prices of resources in the edge-cloud network are adjusted according to changes in supply-demand relationship, and whether to admit the request of each user is decided. Finally, the dynamic distributed inference scheduling decisions are generated through the primal-dual algorithm to maximize the social welfare. Theoretical analysis shows the good competitive ratio and polynomial time complexity of CADEC. Results of simulation experiments present that CADEC improves social welfare by up to 46% compared with state-of-the-art distributed DNN inference schemes. © 2002-2012 IEEE.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Deformable Mixer Transformer with Gating for Multi-Task Learning of Dense Prediction

arXiv

引用

arXiv 2023年

作者： Xu, Yangyang Yang, Yibo Ghanem, Bernard Zhang, Lefei Du, Bo Tao, Dacheng School of Computer Science Wuhan University Wuhan China Visual Computing Center King Abdullah University of Science and Technology Jeddah Saudi Arabia Sydney Ai Centre School of Computer Science The University of Sydney Sydney Australia

Convolution neural networks (CNNs) and Transformers have their own advantages and both have been widely used for dense prediction in multi-task learning (MTL). Most of the current studies on MTL solely rely on CNN or Transformer. In this work, we present a novel MTL model by combining both merits of deformable CNN and query-based Transformer with shared gating for multi-task learning of dense prediction. This combination may offer a simple and efficient solution owing to its powerful and flexible task-specific learning and advantages of lower cost, less complexity and smaller parameters than the traditional MTL methods. We introduce deformable mixer Transformer with gating (DeMTG), a simple and effective encoder-decoder architecture up-to-date that incorporates the convolution and attention mechanism in a unified network for MTL. It is exquisitely designed to use advantages of each block, and provide deformable and comprehensive features for all tasks from local and global perspective. First, the deformable mixer encoder contains two types of operators: the channel-aware mixing operator leveraged to allow communication among different channels, and the spatial-aware deformable operator with deformable convolution applied to efficiently sample more informative spatial locations. Second, the taskaware gating transformer decoder is used to perform the task-specific predictions, in which task interaction block integrated with self-attention is applied to capture task interaction features, and the task query block integrated with gating attention is leveraged to select corresponding task-specific features. Further, the experiment results demonstrate that the proposed DeMTG uses fewer GFLOPs and significantly outperforms current Transformer-based and CNN-based competitive models on a variety of metrics on three dense prediction datasets (i.e., NYUD-v2, PASCAL-Context, and Cityscapes). For example, by using Swin-L as a backbone, our method achieves 57.55 mIoU segmentation

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

DLPO : Towards a Robust, Efficient, and Generalizable Prompt Optimization Framework from a Deep-Learning Perspective

arXiv

引用

arXiv 2025年

作者： Peng, Dengyun Zhou, Yuhang Chen, Qiguang Liu, Jinhao Chen, Jingjing Qin, Libo Research Center for Social Computing and Interactive Robotics Harbin Institute of Technology China School of Computer Science and Engineering Central South University China Fudan University China

Large Language Models (LLMs) have achieved remarkable success across diverse tasks, largely driven by well-designed prompts. However, crafting and selecting such prompts often requires considerable human effort, significantly limiting its scalability. To mitigate this, recent studies have explored automated prompt optimization as a promising solution. Despite these efforts, existing methods still face critical challenges in robustness, efficiency, and generalization. To systematically address these challenges, we first conduct an empirical analysis to identify the limitations of current reflection-based prompt optimization paradigm. Building on these insights, we propose 7 innovative approaches inspired by traditional deep learning paradigms for prompt optimization (DLPO), seamlessly integrating these concepts into text-based gradient optimization. Through these advancements, we progressively tackle the aforementioned challenges and validate our methods through extensive experimentation. We hope our study not only provides valuable guidance for future research but also offers a comprehensive understanding of the challenges and potential solutions in prompt optimization. Our code is available at https://***/sfasfaffa/DLPO. Copyright © 2025, The Authors. All rights reserved.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Simulation-Free Hierarchical Latent Policy Planning for Proactive Dialogues 39

Simulation-Free Hierarchical Latent Policy Planning for Proa...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： He, Tao Liao, Lizi Cao, Yixin Liu, Yuanxing Sun, Yiheng Chen, Zerui Liu, Ming Qin, Bing Research Center for Social Computing and Information Retrieval Harbin Institute of Technology Harbin China Singapore Management University Singapore School of Computer Science Fudan University China

ISBN: (纸本)157735897X

Recent advancements in proactive dialogues have garnered significant attention, particularly for more complex objectives (e.g. emotion support and persuasion). Unlike traditional task-oriented dialogues, proactive dialogues demand advanced policy planning and adaptability, requiring rich scenarios and comprehensive policy repositories to develop such systems. However, existing approaches tend to rely on Large Language Models (LLMs) for user simulation and online learning, leading to biases that diverge from realistic scenarios and result in suboptimal efficiency. Moreover, these methods depend on manually defined, context-independent, coarse-grained policies, which not only incur high expert costs but also raise concerns regarding their completeness. In our work, we highlight the potential for automatically discovering policies directly from raw, real-world dialogue records. To this end, we introduce a novel dialogue policy planning framework, LDPP. It fully automates the process from mining policies in dialogue records to learning policy planning. Specifically, we employ a variant of the Variational Autoencoder to discover fine-grained policies represented as latent vectors. After automatically annotating the data with these latent policy labels, we propose an Offline Hierarchical Reinforcement Learning (RL) algorithm in the latent space to develop effective policy planning capabilities. Our experiments demonstrate that LDPP outperforms existing methods on two proactive scenarios, even surpassing ChatGPT with only a 1.8-billion-parameter LLM. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Epithelial Layer Estimation Using Curvatures and Textural Features for Dysplastic Tissue Detection

引用

computers, Materials & Continua 2021年第4期67卷 761-777页

作者： Afzan Adam Abdul Hadi Abd Rahman Nor Samsiah Sani Zaid Abdi Alkareem Alyessari Nur Jumaadzan Zaleha Mamat Basela Hasan Faculty of Information Science and Technology Center for Artificial Intelligence TechnologyUniversiti Kebangsaan MalaysiaBangiSelangor43300Malaysia Faculty of Science and Technology School of Mathematical SciencesUniversiti Kebangsaan MalaysiaBangiSelangor43300Malaysia Department of Information Technology and Computing Faculty of Computer StudiesArab Open UniversityAmmanJordan

Boundary effect in digital pathology is a phenomenon where the tissue shapes of biopsy samples get distorted during the sampling *** morphological pattern of an epithelial layer is greatly ***,the shape deformation model can normalise the distortions,but it needs a 2D *** theory,on the other hand,is not yet tested on digital pathology ***,this work proposed a curvature detection to reduce the boundary effects and estimates the epithelial *** boundary effect on the tissue surfaces is normalised using the frequency of a curve deviates from being a straight *** epithelial layer’s depth is estimated from the tissue edges and the connected nucleolus ***,the textural and spatial features along the estimated layer are used for dysplastic tissue *** proposed method achieved better performance compared to the whole tissue regions in terms of detecting dysplastic *** result shows a leap of kappa points from fair to a substantial agreement with the expert’s ground truth *** improved results demonstrate that curvatures have been effective in reducing the boundary effects on the epithelial layer of ***,quantifying and classifying the morphological patterns for dysplasia can be *** textural and spatial features on the detected epithelial layer can capture the changes in tissue.

关键词： Digital pathology grading dysplasia tissue boundary effect

来源：评论

学校读者我要写书评

暂无评论

Owl: Differential-Based Side-Channel Leakage Detection for CUDA Applications

Owl: Differential-Based Side-Channel Leakage Detection for C...

引用

International Conference on Dependable Systems and Networks (DSN)

作者： Yu Zhao Wenjie Xue Weijie Chen Weizhong Qiang Deqing Zou Hai Jin National Engineering Research Center for Big Data Technology and System Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China Jinyinhu Laboratory Wuhan China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9798350341058

ISBN: (纸本)9798350341065

Over the past decade, various methods for detecting side-channel leakage have been proposed and proven to be effective against CPU side-channel attacks. These methods are valuable in assisting developers to identify and patch side-channel vulnerabilities. Nevertheless, recent research has revealed the feasibility of exploiting side-channel vulnerabilities to steal sensitive information from GPU applications, which are beyond the reach of previous side-channel detection methods. Therefore, in this paper, we conduct an in-depth examination of various GPU features and present Owl, a novel side-channel detection tool targeting CUDA applications on NVIDIA GPUs. Owl is designed to detect and locate side-channel leakage in various types of CUDA applications. When tracking the execution of CUDA applications, we design a hierarchical tracing scheme and extend the A-DCFG (Attributed Dynamic Control Flow Graph) to address the massively parallel execution in CUDA, ensuring Owl's detection scalability. After completing the initial assessment and filtering, we conduct statistical tests on the differences in program traces to determine whether they are indeed caused by input variations, subsequently facilitating the positioning of side-channel leaks. We evaluate Owl's capability to detect side-channel leaks by testing it on Libgpucrypto, PyTorch, and nvJPEG. Meanwhile, we verify that our solution effectively handles a large number of threads. Owl has successfully identified hundreds of leaks within these applications. To the best of our knowledge, we are the first to implement side-channel leakage detection for general CUDA applications.

关键词： Target tracking Statistical analysis Filtering Scalability Graphics processing units Side-channel attacks Feature extraction

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：