检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Zhang, Shenyi Zhai, Yuchen Guo, keyan Hu, Hongxin Guo, Shengnan Fang, Zheng Zhao, Lingchen Shen, Chao Wang, Cong Wang, Qian Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education School of Cyber Science and Engineering Wuhan University China University at Buffalo United States Xi’an Jiaotong University China City University of Hong Kong Hong Kong

Despite the implementation of safety alignment strategies, large language models (LLMs) remain vulnerable to jailbreak attacks, which undermine these safety guardrails and pose significant security threats. Some defenses have been proposed to detect or mitigate jailbreaks, but they are unable to withstand the test of time due to an insufficient understanding of jailbreak mechanisms. In this work, we investigate the mechanisms behind jailbreaks based on the Linear Representation Hypothesis (LRH), which states that neural networks encode high-level concepts as subspaces in their hidden representations. We define the toxic semantics in harmful and jailbreak prompts as toxic concepts and describe the semantics in jailbreak prompts that manipulate LLMs to comply with unsafe requests as jailbreak concepts. Through concept extraction and analysis, we reveal that LLMs can recognize the toxic concepts in both harmful and jailbreak prompts. However, unlike harmful prompts, jailbreak prompts activate the jailbreak concepts and alter the LLM output from rejection to compliance. Building on our analysis, we propose a comprehensive jailbreak defense framework, JBSHIELD, consisting of two key components: jailbreak detection JBSHIELD-D and mitigation JBSHIELD-M. JBSHIELD-D identifies jailbreak prompts by determining whether the input activates both toxic and jailbreak concepts. When a jailbreak prompt is detected, JBSHIELD-M adjusts the hidden representations of the target LLM by enhancing the toxic concept and weakening the jailbreak concept, ensuring LLMs produce safe content. Extensive experiments demonstrate the superior performance of JBSHIELD, achieving an average detection accuracy of 0.95 and reducing the average attack success rate of various jailbreak attacks to 2% from 61% across distinct LLMs. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Optimal Flying Location for Unmanned Aerial Vehicle Equipped with MF-RIS assisted Covert Wireless on Communication

引用

IEEE Transactions on Vehicular Technology 2025年

作者： Wang, Jie Li, Yunxiang Tian, Wen Dai, Yuewei Liu, Yinqiu Niyato, Dusit Nanjing University of Posts and Telecommunications School of Communications and Information Engineering Nanjing China Nanjing University of Information Science and Technology School of Electronic and Information Engineering Key Laboratory of Intelligent Support Technology In Complex Environment Ministry of Education Nanjing210044 China Nanyang Technological University College of Computing and Data Science Singapore

In the current information age, covert communication holds significant value owing to its advantages, such as heightened security, enhanced flexibility, and improved concealment, surpassing those of traditional communication methods. However, existing covert communication approaches approximate communication channel models by Line-of-Sight (LoS) links, which is insufficient due to neglecting the impact of multipath effects and applying inaccurate performance assessments. To better adapt to real-world communication channel conditions, we propose a novel covert communication model assisted by Unmanned Aerial Vehicles (UAV) equipped with Multi-Functional Reconfigurable Intelligent Surface (MF-RIS), named UMCCM. Specifically, based on a mixed link of LoS and Non-Line-of-Sight (NLoS), UMCCM leverages MF-RIS as a flexible relay to enhance the signal-to-noise ratio (SNR) and effective throughput. We first derive the mixed channel model expression. Then, we transform covert constraints from incomplete gamma functions to those linked with relative entropy and accordingly optimize the UAV transmission power and block length. Finally, the optimal flying position of the UAV is obtained by analyzing the functional properties of SNR and effective throughput under these refined constraints. Extensive numerical results demonstrate the superiority of our proposed UMCCM for enhancing SNR and effective throughput over the existing works. © 2025 IEEE.

关键词： Aircraft communication

来源：评论

学校读者我要写书评

暂无评论

Towards Effective Detection of Ponzi Schemes on Ethereum with Contract Runtime Behavior Graph

引用

ACM Transactions on Software engineering and Methodology 2025年第4期34卷

作者： Liang, Ruichao Chen, Jing Wu, Cong He, Kun Wu, Yueming Sun, Weisong Du, Ruiying Zhao, Qingchuan Liu, Yang Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education School of Cyber Science and Engineering Wuhan University Wuhan China Nanyang Technological University Singapore Singapore School of Computer Science and Engineering Nanyang Technological University Singapore Singapore City University of Hong Kong Hong Kong Hong Kong

Ponzi schemes, a form of scam, have been discovered in Ethereum smart contracts in recent years, causing massive financial losses. Existing detection methods primarily focus on rule-based approaches and machine learning techniques that utilize static information as features. However, these methods have significant limitations. Rule-based approaches rely on pre-defined rules with limited capabilities and domain knowledge dependency. Using static information like opcodes for machine learning fails to effectively characterize Ponzi contracts, resulting in poor reliability and interpretability. Our research shows no significant difference between Ponzi and non-Ponzi contracts at the opcode level. Moreover, relying on static information like transactions for machine learning requires a certain number of transactions to achieve detection, which limits the scalability of detection and hinders the identification of 0-day Ponzi schemes. In this article, we propose PonziGuard, an efficient Ponzi scheme detection approach based on contract runtime behavior. Inspired by the observation that a contract's runtime behavior is more effective in disguising Ponzi contracts from the innocent contracts, PonziGuard establishes a comprehensive graph representation called contract runtime behavior graph (CRBG), to accurately depict the behavior of Ponzi contracts. Furthermore, it formulates the detection process as a graph classification task on CRBG, enhancing its overall effectiveness. The experiment results show that PonziGuard surpasses the current state-of-the-art approaches in the ground-truth dataset, achieving a precision of 96.9%, recall of 98.2%, and F1-score of 97.5%. It also exhibits the highest level of interpretability among the current tools. We applied PonziGuard to Ethereum Mainnet and demonstrated its effectiveness in real-world scenarios. Using PonziGuard, we identified 805 Ponzi contracts on Ethereum Mainnet, which have resulted in an estimated economic loss of 281,700

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

TDiffSal: Text-Guided Diffusion Saliency Prediction Model for Images 27th

TDiffSal: Text-Guided Diffusion Saliency Prediction Model fo...

引用

27th International Conference on Pattern Recognition, ICPR 2024

作者： Zhang, Nana Xiong, Min Zhu, Dandan Zhu, Kun Zhai, Guangtao School of Computer Science and Technology Donghua University No. 2999 Renmin North Road Songjiang District Shanghai201620 China State Key Laboratory for Novel Software Technology Nanjing University Nanjing210093 China Institute of AI Education Shanghai East China Normal University No. 3663 Zhongshan North Road Putuo District Shanghai200333 China Key Laboratory of Embedded System and Service Computing Ministry of Education Tongji University No. 4800 Cao’an Highway Jiading District Shanghai201804 China Institute of Image Communication and Network Engineering Shanghai Jiao Tong University No. 800 Dongchuan Road Minhang District Shanghai200240 China

ISBN: (纸本)9783031781858

Existing visual saliency prediction methods mainly focus on single-modal visual saliency prediction, while ignoring the significant impact of text on visual saliency. To more comprehensively explore the influence of text on human attention in images, we propose a text-guided diffusion saliency prediction model, named TDiffSal. In specific, recent studies on stable diffusion models have shown promising performance in unifying tasks due to their inherent generalization ability. Inspired by this, a novel diffusion model for generalized visual-text saliency prediction is proposed, which formulates the prediction issue as a conditional generative task of the saliency map by employing input visual and text as the conditions. Meanwhile, we introduce a multi-head fusion module to effectively integrate text features and image features, which can efficiently guide the image denoising process and progressively refine the generated saliency map to make it semantically relevant to the text. Additionally, we employ an efficient pre-training strategy to enhance the robustness and generalization of the proposed model. We conduct extensive experiments on benchmark datasets to demonstrate its superior performance compared to other state-of-the-art methods. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Modeling of self-pressurization process with low flow supply of carbon dioxide

引用

Zhongnan Daxue Xuebao (Ziran Kexue Ban)/Journal of Central South University (Science and Technology) 2025年第4期56卷 1628-1638页

作者： Xiong, Yu Lu, Zhaijun Gan, Ning School of Transportation Engineering Central South University Changsha 410075 China Center of Advanced Scientific Research on Extreme Fluid Power Central South University Changsha 410075 China Key Laboratory of Traffic Safety on the Track of Ministry of Education Central South University Changsha 410075 China

To study the self-pressurization process with low flow supply of carbon dioxide, a new non-thermal equilibrium self-pressurization mathematical model based on the quality to characterize gasification rate of liquid-phase carbon dioxide was proposed. The carbon dioxide self-pressurization supply test platform was established, and three tests with different initial conditions for low flow supply of carbon dioxide were conducted. The results of the mathematical model calculations were compared with the experimental results. The effects of initial temperature(from 278.15 K to 308.15 K) and filling density (800 kg/m³ to 1000 kg/m³⁾ on the effective supply rate of liquid-phase carbon were studied. The results indicate that the mathematical model calculations of pressure and temperature are in good agreement with the experimental results, validating the accuracy of the model. In the modeling process, the heat exchange between the storage tank and carbon dioxide should be considered, while the heat flux between liquid-phase and gas-phase carbon dioxide is relatively small and can be neglected in the modeling. Higher initial temperature of carbon dioxide and lower filling density can result in higher gasification mass during the carbon dioxide supply process, reducing the effective supply rate. © 2025 Central South University. All rights reserved.

关键词： carbon dioxide low flow supply quality self-pressurization

来源：评论

学校读者我要写书评

暂无评论

DAT: Dual-branch Adapter-Tuning for Few-shot Recognition

引用

IEEE Transactions on Circuits and Systems for Video Technology 2025年

作者： Chen, Junxi Wu, Guangxing Li, Hongxiang Chen, Jiankang Zhang, Wentao Zheng, Weishi Wang, Ruixuan Sun Yat-sen Univerisity College of Computer Science and Engineering Guangzhou510275 China Peking University School of Electronic and Computer Engineering Shenzhen518055 China Ministry of Education Key Laboratory of Machine Intelligence and Advanced Computing Guangzhou510275 China Sun Yat-sen Univerisity School of Computer Science and Engineering Guangzhou510275 China Peng Cheng Laboratory Shenzhen518066 China

Parameter-Efficient Fine-Tuning methods based on vision-language models (such as CLIP) for few-shot learning have recently received considerable attention. However, previous works only fine-tune either the image or text branch, breaking the alignment of the original two branches, meanwhile fine-tuning both branches of the CLIP would inevitably introduce more trainable parameters and likely cause more severe over-fitting due to the limited training data. In this study, we propose a novel Dual-branch Adapter-Tuning framework (DAT), which collaboratively trains the visual adapter and textual adapter added to the two branches of the original CLIP with multiple consistency constraints. By effectively utilizing the semantically detailed class-specific prompts and outputs of the original CLIP to guide the fine-tuning of both branches, our method gains exceptional adaptation ability to the downstream few-shot learning tasks and alleviates the over-fitting issue, meanwhile maximally preserving the generalization ability of the original CLIP model. Our proposed framework has achieved superior performance on diverse datasets under various few-shot learning settings compared to the existing approaches. The source code is available at https://***/SandyXi/DAT. © 1991-2012 IEEE.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

Incorporating edge sharpening and covariance attention for named entity recognition

引用

Neurocomputing 2025年 643卷

作者： Yang, Caiwei Chen, Yanping Yu, Shuai Huang, Ruizhang Qin, Yongbin Engineering Research Center of Text Computing & Cognitive Intelligence Ministry of Education Guizhou University Guizhou Guiyang550025 China State Key Laboratory of Public Big Data Guizhou university Guizhou Guiyang550025 China College of Computer Science and Technology Guizhou University Guiyang550025 China

Named Entity Recognition (NER) is a key application in the field of Artificial Intelligence and Natural Language Processing, which automatically identifies and categorizes entities in text by intelligent algorithms. In NER, all spans of a sentence can be organized into a two-dimensional representation. The semantic plane has the advantage to represent the semantic structure of a sentence and to learn the interaction between spans. One of an important phenomenon of this representation is that neighboring elements of the semantic plane are spans denoted to overlapped subsequences in a sentence. Because they share the same contextual features and semantic dependencies, it is difficult to distinguish true entities from the backgrounds. Therefore, refining span representations and building the semantic dependency between spans is helpful for the entity recognition task. In this paper, we propose an Edge Sharpening and Covariance Attention (ES&CA) model to support recognizing named entities from the semantic plane representation. The edge sharpening (ES) module adopts a differential convolution to sharpen the semantic gradients in the semantic plane, which has the ability to gather semantic information from neighborhoods. In the covariance attention (CA) module, the covariance between spans are applied to weight the attention of spans relevant to task-relevant learning objective. Establishing semantic relationships across spans is a highly successful approach. The ES&CA model is assessed on five public datasets for nested and flattened named entity recognition. The evaluation results demonstrate the effectiveness of our strategy in distinguishing entity spans from the backgrounds, hence significantly enhancing the final performance. © 2025 Elsevier B.V.

关键词： Natural language processing systems

来源：评论

学校读者我要写书评

暂无评论

Subband Adaptive Filtering Based on the Squared Sine Error Criterion: Analysis and Applications

引用

IEEE Transactions on Audio, Speech and Language Processing 2025年 33卷 1949-1959页

作者： Zhikai Gong Yingsong Li Xinqi Huang Liping Li Zhixiang Huang Rodrigo. C. de Lamare Yingying Zhu College of Information and Communication Engineering Harbin Engineering University Harbin China Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University Hefei China Centre for Telecommunications Research (CETUC) Pontifical Catholic University of Rio de Janeiro (PUC-Rio) Gávea Brazil School of Physics Engineering and Technology University of York York U.K.

In this work, we present a subband adaptive filtering (SAF) algorithm with good robustness for impulsive noise environments. The proposed squared sine normalized SAF (SS-NSAF) algorithm employs a squared sine function and is inspired by the idea of normalization. A statistical analysis of the proposed SS-NSAF algorithm is carried out under impulsive noise environments, which includes stability conditions for the convergence and analytical formulas to predict the mean-square deviation of the SS-NSAF algorithm. Numerical simulations for system-identification and echo cancellation applications are conducted to illustrate the performance of the SS-NSAF algorithm and competing techniques in impulsive noise environments.

关键词： Noise Prediction algorithms Signal processing algorithms Filtering algorithms Nonlinear filters Maximum likelihood detection Convergence Vectors Mathematical models Information filters

来源：评论

学校读者我要写书评

暂无评论

CMCache: An Adaptive Cross-Level Data Placement Method for Multi-Level Cache

引用

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems 2025年

作者： Zeng, Zhaoyang Tan, Yujuan Ma, Zhulin Li, Jiali Zhao, Sanle Liu, Duo Chen, Xianzhang Ren, Ao Chongqing University College of Computer Science Chongqing400044 China Chongqing University Key Laboratory of Dependable Service Computing in Cyber Physical Society Ministry of Education Chongqing400044 China Chongqing Universityof Posts and Telecommunications College of Software Engineering China Chongqing University College of Big Data and Software Engineering Chongqing China

Multi-level cache systems enhance I/O performance by optimizing data placement across various cache levels from a global perspective. However, existing methods often struggle to place data at the optimal cache level promptly due to their reliance on historical access patterns and inflexible placement strategies. These methods face two main challenges: (1) For already cached data with sufficient access history, existing approaches only optimize movement between adjacent cache levels, potentially delaying data arrival at its globally optimal cache level and leading to unnecessary bandwidth consumption and increased latency. (2) For newly entered data without access history, current methods cannot accurately predict their future hotness and simply place them at a fixed cache level (i.e. first or last level), overlooking future accesses of new data and potentially resulting in high cache miss rates or cache pollution. To address these issues, we propose CMCache, an adaptive cross-level data placement method for multi-level cache. CMCache applies distinct placement strategies for cached and new data to reach the optimal level timely, considering their different characteristics. It also logically divides cache space into two sections to manage cached and new data separately, dynamically adjusting section sizes based on access patterns. This approach significantly improves data placement efficiency, achieving up to an 89% reduction in miss rates and a 79% decrease in average response times compared to existing methods. © 2025 IEEE.

关键词： Cache memory

来源：评论

学校读者我要写书评

暂无评论

MODFuzz: A Multi-Objective Directed Fuzzer for USB Drivers

引用

IEEE Internet of Things Journal 2025年

作者： Zhou, Yilin Peng, Guojun Wang, Xingliang Wang, Chenyang Li, Zichuan Liu, Side Wang, Yanhao Yang, Xiuzhang Fu, Jianming Wuhan University Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education School of Cyber Science and Engineering Wuhan430072 China Indiana University Bloomington BloomingtonIN United States Nio Inc Beijing China Guizhou University Guizhou Big Data Academy Guiyang China

USB interfaces have become ubiquitous in various Internet of Things (IoT) devices, all adhering to the same USB protocol. While enhancing convenience, they also widen the potential attack surface. Fuzzing is a proactive way to identify potential security threats for USB drivers. However, existing USB driver fuzzers primarily prioritize the code coverage of USB drivers, leading to a significant waste of computational resources on irrelevant code segments. To this end, we combine directed fuzzing and USB driver fuzzing for the first time, and present MODFuzz, a pioneering multi-objective directed fuzzing method for USB drivers. MODFuzz autonomously locates the most vulnerable parts within USB drivers, concentrating fuzzing efforts on these areas. Diverging from the existing directed fuzzers, MODFuzz employs a dynamic direction instead of predetermined addresses to guide the fuzzing campaign toward the triggered execution traces with a greater probability of containing vulnerabilities. MODFuzz outperforms the strong baseline in terms of execution speed (about 14% improvement) and crash generation capabilities (about 69% improvement). Meanwhile, we found six previously unknown bugs (all confirmed and assigned vulnerability IDs) in Linux kernel v6.4.10 and received acknowledgment from Red Hat. © 2025 IEEE.

关键词： Linux

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：