检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Yuan, Yu Zhao, Lili Zhang, Kai Zheng, Guangting Liu, Qi State Key Lab of Cognitive Intelligence University of Science and Technology of China China School of Computer Science and Technology University of Science and Technology of China China

Large Language Models (LLMs) have shown remarkable capabilities in various natural language processing tasks. However, LLMs may rely on dataset biases as shortcuts for prediction, which can significantly impair their robustness and generalization capabilities. This paper presents Shortcut Suite, a comprehensive test suite designed to evaluate the impact of shortcuts on LLMs’ performance, incorporating six shortcut types, five evaluation metrics, and four prompting strategies. Our extensive experiments yield several key findings: 1) LLMs demonstrate varying reliance on shortcuts for downstream tasks, significantly impairing their performance. 2) Larger LLMs are more likely to utilize shortcuts under zero-shot and few-shot in-context learning prompts. 3) Chain-of-thought prompting notably reduces shortcut reliance and outperforms other prompting strategies, while few-shot prompts generally underperform compared to zero-shot prompts. 4) LLMs often exhibit overconfidence in their predictions, especially when dealing with datasets that contain shortcuts. 5) LLMs generally have a lower explanation quality in shortcut-laden datasets, with errors falling into three types: distraction, disguised comprehension, and logical fallacy. Our findings offer new insights for evaluating robustness and generalization in LLMs and suggest potential directions for mitigating the reliance on shortcuts. The code is available at https://***/yyhappier/***. Copyright © 2024, The Authors. All rights reserved.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Streamlining Computational Lithography With Efficient Pattern Database

引用

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

作者： Zheng, Su Zhao, Wenqian Sun, Shuyuan Yang, Fan Yu, Bei Wong, Martin D.F. The Chinese University of Hong Kong Department of Computer Science and Engineering Hong Kong Fudan University State Key Lab of Integrated Circuits and Systems China Hong Kong Baptist University Department of Computer Science Hong Kong

In the pursuit of advancing computational lithography, this paper introduces a novel pattern database framework designed to support related tasks. The proposed framework is built upon three core components: an unsupervised metric learning method for robust pattern embedding, a vector database for swift pattern retrieval, and an efficient algorithm dedicated to pattern clustering. These elements synergize to significantly enhance the efficiency and effectiveness of various computational lithography methods. In downstream tasks, our framework provides accurate lithography hotspot detection through pattern retrieval, streamlines inverse lithography technique (ILT) by leveraging solution reusing, and facilitates the exploration of ILT & source parameters based on the pattern clustering results. Collectively, these advancements culminate in a comprehensive improvement in computational lithography, offering a scalable solution for the ever-evolving demands of this field. © 2025 IEEE.

关键词： Inverse problems

来源：评论

学校读者我要写书评

暂无评论

Reach-avoid Controllers Synthesis for Safety Critical Systems

arXiv

引用

arXiv 2023年

作者： Xue, Bai State Key Lab. of Computer Science Institute of Software Cas Beijing China

In this paper we propose sufficient conditions to synthesizing reach-avoid controllers for deterministic systems modelled by ordinary differential equations and stochastic systems modeled by stochastic differential equations based on the notion of control guidance-barrier functions. We begin with considering deterministic systems. Given an open safe set, a target set and a nominal controller, we attempt to synthesize a reach-avoid controller, which modifies the nominal controller in a minimal way enforcing the reach-avoid objective, i.e., the system will enter the target set eventually while staying inside the safe set before the first target hitting time. Three control guidance-barrier functions are developed and three corresponding conditions for synthesizing reach-avoid controllers are constructed, which get progressively weaker and thus facilitate the optimal controller design. The first and second ones are termed exponential and asymptotic control guidance-barrier functions, which guarantee that every trajectory starting from the safe set will enter the target set respectively at an exponential rate and in an asymptotic way. However, it is observed that these two control guidance-barrier functions have one condition in common of enforcing invariance of its every positive sublevel set until the system enters the target set, which is strict and thus limits the space of admissible controllers. Consequently, a lax control guidance-barrier function is further developed such that only the safe set set is an invariance before the system enters the target set, expanding the space of admissible control inputs. Then, we extend the methodologies for deterministic systems to stochastic systems, which synthesize reach-avoid controllers in the probabilistic setting. Finally, several numerical examples demonstrate the performance of proposed methods. Copyright © 2023, The Authors. All rights reserved.

关键词： Controllers

来源：评论

学校读者我要写书评

暂无评论

Learning Inverse Dynamics of Multi-segmented Larval Crawling Neuromechanics 10

Learning Inverse Dynamics of Multi-segmented Larval Crawling...

引用

10th IEEE Smart World Congress, SWC 2024

作者： Tang, Weihao Zhang, Hao Gong, Zhefeng Zheng, Nenggan Zhejiang University Qiushi Academy for Advanced Studies College of Computer Science and Technology Hangzhou China Zhejiang University School of Medicine School of Brain Science and Brain Medicine Department of Neurology of the Fourth Hospital Hangzhou310058 China Zhejiang University Qiushi Academy for Advanced Studies The State Key Lab of Brain-Machine Intelligence College of Computer Science and Technology Hangzhou China Bengbu University Bengbu China

ISBN: (纸本)9798331520861

Neuromechanical modeling that bridges neural activity and locomotion patterns is essential for studying locomotion control. The Drosophila larva, with its diverse softbodied locomotion, serves as an ideal subject for biomimetic research. Constructing a neuromechanical model of Drosophila larvae presents two primary challenges: efficiently simulating the larva's soft, multi-segmented body and controlling locomotion within a large action space by actuating muscle groups. This work presents a three-dimensional neuromechanical model of the larva and introduces an inverse dynamics learning method for locomotion control. Our approach efficiently simulates the larva's multi-segmented, deformable body by integrating computational modules for soft tissues, muscles, and friction. We employ a parameterized controller to reproduce peristaltic waves and use a hybrid artificial neural network architecture to learn inverse dynamics from partially observed states. Balancing biological interpretability with modeling complexity, this model advances the understanding of brain-body-environment interactions in Drosophila larvae, contributing to nature-inspired intelligent systems research. © 2024 IEEE.

关键词： Intelligent systems

来源：评论

学校读者我要写书评

暂无评论

Sparse Attention-based Hierarchical Node Representation for Spatial Domain Identification

Sparse Attention-based Hierarchical Node Representation for ...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Peng, Wei Ping, Zhihao Dai, Wei Fu, Xiaodong Liu, Li Liu, Lijun Yu, Ning Kunming University of Science and Technology Faculty of Information Engineering and Automation Kunming China Computer Technology Application Key Lab of Yunnan Province Kunming University of Science and Technology Kunming China State University of New York The College at Brockport Department of Computing Sciences BrockportNY14422 United States

ISBN: (纸本)9798350386226

Using deep learning models on spatial transcriptomics data to identify the spatial domain is crucial for uncovering the spatial distribution of cells and gene expression patterns within tissues, essential for understanding complex biological processes and disease mechanisms. Existing methods for spatial domain partitioning often rely on predefined adjacency relationships at a single scale, overlooking the hierarchical structure and functional characteristics of biological tissues. In this paper, we propose SpaNFM, a novel method that leverages sparse attention-based hierarchical node representation and multi-view contrastive learning for spatial domain identification in spatial transcriptomics data. The SpaNFM first treats each spot as a node and constructs two views using different data augmentation techniques based on tissue image information, gene expression profiles, and spatial coordinates of cells. Subsequently, SpaNFM utilizes a sparse attention-based hierarchical node fusion module to generate coarse-grained node representations. This fine-to-coarse hierarchical structure integrates complementary information from multi-granularity node features and reduces model complexity due to the decreased node size. The model parameters are updated using gene expression reconstruction loss and contrastive loss on the coarse-grained node representations from the two views. Finally, the learned node features are subjected to downstream clustering using the Leiden algorithm. We tested SpaNFM on the human dorsolateral prefrontal cortex dataset. The results demonstrate that SpaNFM outperforms other state-of-the-art methods in most cases. The data and code are available at: https://***/weiba/SpaNFM © 2024 IEEE.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Bio-assembled smart nanocapsules for targeted delivery of KRAS shRNA and cancer cell bioimage

引用

Chinese Chemical Letters 2023年第1期34卷 270-273页

作者： Maonan Wang Zengchao Guo Jiayu Zeng Liu Liu Yihan Wang Jinpeng Wang Hongbing Lu Haijun Zhang Hui Jiang Xuemei Wang State Key Laboratory of Bioelectronics(Chien-Shiung Wu Lab) School of Biological Science and Medical EngineeringSoutheast UniversityNanjing 210096China Department of Biomedical Engineering/Computer Application Fourth Military Medical UniversityXi’an 710032China Department of Oncology Zhongda HospitalMedical School of Southeast UniversityNanjing 210009China

The five-year survival rate for pancreatic cancer is less than 5%. However, the current clinical multimodal therapy combined with first-line chemotherapy drugs only increases the patient’s median survival from 5.0 months to 7.2 months. Consequently, a new strategy of cancer treatments is urgently needed to overcome this high-fatality disease. Through a series of biometric analyses, we found that KRAS is highly expressed in the tumor of pancreatic cancer patients, and this high expression is closely related to the poor prognosis of patients. It shows that inhibiting the expression of KRAS has great potential in gene therapy for pancreatic cancer. Given those above, we have exploited the possibility of targeted delivery of KRAS shRNA with the intelligent and bio-responsive nanomedicine to detect the special oxidative stress microenvironment of cancer cells and realize efficient cancer theranostics. Our observations demonstrate that by designing the smart self-assembled nanocapsules of melanin with fluorescent nanoclusters we can readily achieve the bio-recognition and bioimaging of cancer cells in biological solution or *** self-assembled nanocapsules can make a significant bio-response to the oxidative stress microenvironment of cancer cells and generate fluorescent zinc oxide Nanoclusters in situ for targeted cell bioimaging. Moreover, it can also readily facilitate cancer cell suppression through the targeted delivery of KRAS shRNA and low-temperature hyperthermia. This raises the possibility to provide a promising theranostics platform and self-assembled nanomedicine for targeted cancer diagnostics and treatments through special oxidative stress-responsive effects of cancer cells.

关键词： Smart bio-responsive nanocapsules Intelligent nanomedicine shRNA KRAS Fluorescent bioimaging Oxidative stress response

来源：评论

学校读者我要写书评

暂无评论

UNIVERSAL VISION-LANGUAGE DENSE RETRIEVAL: LEARNING A UNIFIED REPRESENTATION SPACE FOR MULTI-MODAL RETRIEVAL 11

UNIVERSAL VISION-LANGUAGE DENSE RETRIEVAL: LEARNING A UNIFIE...

引用

11th International Conference on Learning Representations, ICLR 2023

作者： Liu, Zhenghao Xiong, Chenyan Lv, Yuanhuiyi Liu, Zhiyuan Yu, Ge Department of Computer Science and Technology Northeastern University Shenyang China Microsoft Research Redmond United States Department of Computer Science and Technology Tsinghua University Beijing China Institute for Artificial Intelligence Tsinghua University Beijing China State Key Lab on Intelligent Technology and Systems Tsinghua University Beijing China

This paper presents Universal Vision-Language Dense Retrieval (UniVL-DR), which builds a unified model for multi-modal retrieval. UniVL-DR encodes queries and multi-modality resources in an embedding space for searching candidates from different modalities. To learn a unified embedding space for multi-modal retrieval, UniVL-DR proposes two techniques: 1) Universal embedding optimization strategy, which contrastively optimizes the embedding space using the modality-balanced hard negatives;2) Image verbalization method, which bridges the modality gap between images and texts in the raw data space. UniVL-DR achieves the state-of-the-art on the multi-modal open-domain question answering benchmark, WebQA, and outperforms all retrieval models on the two subtasks, text-text retrieval and text-image retrieval. It demonstrates that universal multi-modal search is feasible to replace the divide-and-conquer pipeline with a united model and also benefits single/cross modality tasks. All source codes of this work are available at https://***/OpenMatch/UniVL-DR. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

A Dynamic Load Balancing Algorithm for Stable Pushing of Docker Images in Nginx 6

A Dynamic Load Balancing Algorithm for Stable Pushing of Doc...

引用

6th International Conference on Machine Learning, Big Data and Business Intelligence, MLBDBI 2024

作者： Xi, Hajie Hu, Song Zhang, He Zhang, Shuang Weng, Jingnong Huang, Jian School of Computer Science and Engineering Beihang University Beijing China Research Institute of Highway Ministry of Transport State Key Lab of Transportation System Beijing China Transportation Comprehensive Enforcement Corps Beijing China School of Software Beihang University Beijing China

ISBN: (纸本)9798331541798

With the rapid development of artificial intelligence, a wide variety of deep learning models have emerged across various fields. Nevertheless, the quality of these models varies significantly, making the effective evaluation of model performance a pressing issue. In some open data competitions, remote evaluation based on Docker images has increasingly become a novel and reliable approach, providing a consistent environment to ensure the fairness of model evaluations. However, it imposes high demands on hardware resources, typically requiring dedicated resources for each participant to serve as Docker Registries for the pushing and management of Docker images. In this paper, we extend the load balancing module of the Nginx server by integrating Lua scripts and propose a Cache-based Dynamic Smooth Weighted Round-robin (CDSWRR) algorithm to optimize the pushing process of Docker images. By constructing a unified load balancing entry, the proposed approach simplifies the deployment of Docker Registry clusters. Experiments show that compared to weighted round-robin and IP hash algorithms of Nginx, the CDSWRR algorithm is able to ensure stable pushing of Docker images and reduce both response error rate and response time. © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

CFSynthesis: Controllable and Free-view 3D Human Video Synthesis

arXiv

引用

arXiv 2024年

作者： Cui, Liyuan Xui, Xiaogang Dong, Wenqi Yang, Zesong Bao, Hujun Cui, Zhaopeng State Key lab of CAD&CG College of Computer Science Zhejiang University China Department of computer science and engineering the Chinese University of Hong Kong Hong Kong

Human video synthesis aims to create lifelike characters in various environments, with wide applications in VR, storytelling, and content creation. While 2D diffusion-based methods have made significant progress, they struggle to generalize to complex 3D poses and varying scene backgrounds. To address these limitations, we introduce CFSynthesis, a novel framework for generating high-quality human videos with customizable attributes, including identity, motion, and scene configurations. Our method leverages a texture-SMPL-based representation to ensure consistent and stable character appearances across free viewpoints. Additionally, we introduce a novel foregroundbackground separation strategy that effectively decomposes the scene as foreground and background, enabling seamless integration of user-defined backgrounds. Experimental results on multiple datasets show that CFSynthesis not only achieves stateof- the-art performance in complex human animations but also adapts effectively to 3D motions in free-view and user-specified scenarios. Copyright © 2024, The Authors. All rights reserved.

关键词： Animation

来源：评论

学校读者我要写书评

暂无评论

Disentangled latent factors for muti-cause treatment effect estimation 6

Disentangled latent factors for muti-cause treatment effect ...

引用

2024 6th International Conference on Artificial Intelligence Technologies and Applications, ICAITA 2024

作者： Zhao, Zhuoyue Wang, Haotian Xu, Liyang Wang, Qian Tang, Yuhua Department of Intelligent Data Science National University of Defense Technology Changsha410073 China Intelligent Game and Decision Lab. Academy of Military Science Beijing100071 China Institute for Quantum Information State Key Laboratory of High-Performance Computing College of Computer Science and Technology National University of Defense Technology Changsha410073 China

Existing methods estimate treatment effects from observational data and assume that covariates are all confounders. However, observed covariates may not directly represent confounding variables that influence both treatment and outcome. They always include variables that only affect the treatment or the outcome. In addition, for multi-dimensional binary treatments, disentangled methods are mainly designed for binary variables and ignore the impact of multi-cause treatment variables on the inference of latent factors. To address these two issues, based on variable decomposition and proxy inference, we propose the Disentangled Latent Factors for Multi-cause Treatment Estimation (DEMTE) algorithm. It utilizes an identifiable autoencoder to infer and disentangle latent factors based on the joint distribution of variables in observational data. DEMTE evaluates the treatment effect on the disentangled factors. Synthetic experiments and semi-synthetic experiments demonstrate the effectiveness of the inference and disentanglement techniques and our method achieves more accurate treatment effect estimation. © 2024 Institute of Physics Publishing. All rights reserved.

关键词：

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：