Found a total of 10000 related content
ICML 2024 | Revealing the mechanism of non-linear Transformer learning and generalization in contextual learning
Article Introduction:The AIxiv column is a column where this site publishes academic and technical content. In the past few years, the AIxiv column of this site has received more than 2,000 reports, covering top laboratories from major universities and companies around the world, effectively promoting academic exchanges and dissemination. If you have excellent work that you want to share, please feel free to contribute or contact us for reporting. Submission email: liyazhou@jiqizhixin.com; zhaoyunfeng@jiqizhixin.com The author of this article, Li Hongkang, is a doctoral candidate in the Department of Electrical, Computer and Systems Engineering at Rensselaer Polytechnic Institute in the United States. He graduated from the University of Science and Technology of China with a bachelor's degree. Research directions include deep learning theory, large language model theory, statistical machine learning, etc. Currently at ICLR/
2024-06-29
comment 0
433
Give the RAG system a comprehensive 'physical examination' with Amazon's open source RAGChecker diagnostic tool
Article Introduction:The AIxiv column is a column where this site publishes academic and technical content. In the past few years, the AIxiv column of this site has received more than 2,000 reports, covering top laboratories from major universities and companies around the world, effectively promoting academic exchanges and dissemination. If you have excellent work that you want to share, please feel free to contribute or contact us for reporting. Submission email: liyazhou@jiqizhixin.com; zhaoyunfeng@jiqizhixin.com Amazon Shanghai Artificial Intelligence Research Institute was established in 2018 and has become one of the leading institutions in the field of deep learning research, publishing a total of ~90 papers. Research areas include basic theories of deep learning, natural language processing, computer vision, graph machine learning, high performance
2024-08-19
comment 0
887
Integrating more than 200 related studies, the latest review of the large model 'lifelong learning' is here
Article Introduction:The AIxiv column is a column where this site publishes academic and technical content. In the past few years, the AIxiv column of this site has received more than 2,000 reports, covering top laboratories from major universities and companies around the world, effectively promoting academic exchanges and dissemination. If you have excellent work that you want to share, please feel free to contribute or contact us for reporting. Submission email: liyazhou@jiqizhixin.com; zhaoyunfeng@jiqizhixin.com The authors of this paper are all from the team of Professor Ma Qianli of South China University of Technology, and their laboratory is the Machine Learning and Data Mining Laboratory. The three co-first authors of the paper are doctoral student Zheng Junhao, master's student Qiu Shengjie, and master's student Shi Chengming. Their main research directions include large models and final models.
2024-09-02
comment 0
271
Integrating more than 200 related studies, the latest review of the large model 'lifelong learning' is here
Article Introduction:The AIxiv column is a column where this site publishes academic and technical content. In the past few years, the AIxiv column of this site has received more than 2,000 reports, covering top laboratories from major universities and companies around the world, effectively promoting academic exchanges and dissemination. If you have excellent work that you want to share, please feel free to contribute or contact us for reporting. Submission email: liyazhou@jiqizhixin.com; zhaoyunfeng@jiqizhixin.com The authors of this paper are all from the team of Professor Ma Qianli of South China University of Technology, and their laboratory is the Machine Learning and Data Mining Laboratory. The three co-first authors of the paper are doctoral student Zheng Junhao, master's student Qiu Shengjie, and master's student Shi Chengming. Their main research directions include large models and final models.
2024-09-02
comment 0
982
Introducing OverflowAI, Stack Overflow adds AI technology to improve product capabilities
Article Introduction:StackOverflowStackOverflow is a knowledge sharing platform and question and answer community for programmers founded in 2008. It covers a wide range of topics including programming, software development, algorithms, data structures, operating systems, databases, networking, and more. On StackOverflow, programmers can ask questions, answer questions, share experiences and knowledge, and participate in discussions. Through its unique Q&A mechanism and community-driven content contribution model, StackOverflow has become one of the largest programming Q&A websites in the world. Its user base includes professional developers, students, researchers and enthusiasts in various technical fields. Whether you are a beginner or a seasoned expert, you can
2023-09-01
comment 0
917
Use AI assistants to optimize the quality and efficiency of legal professional paper writing
Article Introduction:In thesis writing of law majors, the Manujian AI assistant can become a powerful tool to improve the quality and efficiency of writing. This article will explore how to use AI assistants to improve the quality and efficiency of writing professional legal papers. The contents that need to be rewritten by the Draft View AI Assistant are: 1. Automatically retrieve and organize legal information. The Draft View AI Assistant can help students quickly obtain legal cases, academic articles and legal information through its efficient search function. It can organize and filter this information, provide students with a clear research framework, avoid tedious manual searching and sorting, and improve writing efficiency. The purpose of providing structured writing frameworks and templates for the draft AI assistant is to help learners better organize and express one's thoughts. By using these frameworks and templates, learners can more easily
2023-09-13
comment 0
911
Is a ghost controlling your phone? Large model GUI agents are vulnerable to environment hijacking
Article Introduction:The AIxiv column is a column where this site publishes academic and technical content. In the past few years, the AIxiv column of this site has received more than 2,000 reports, covering top laboratories from major universities and companies around the world, effectively promoting academic exchanges and dissemination. If you have excellent work that you want to share, please feel free to contribute or contact us for reporting. Submission email: liyazhou@jiqizhixin.com; zhaoyunfeng@jiqizhixin.com The first author of this article, Ma Xinbei, is a fourth-year doctoral student in the Department of Computer Science at Shanghai Jiao Tong University. His research interests include autonomous agents, reasoning, and the interpretability and knowledge of large models. edit. The work was jointly completed by Shanghai Jiao Tong University and Meta. Thesis title: Cautio
2024-09-02
comment 0
319
The black box has been opened! Transformer visual explanation tool that can be played, runs GPT-2 locally, and can also perform real-time reasoning
Article Introduction:It's 2024, is there anyone who still doesn't understand how Transformer works? Come and try this interactive tool. In 2017, Google proposed Transformer in the paper "Attentionisallyouneed", which became a major breakthrough in the field of deep learning. The number of citations of this paper has reached nearly 130,000. All subsequent models of the GPT family are also based on the Transformer architecture, which shows its wide influence. As a neural network architecture, Transformer is popular in a variety of tasks from text to vision, especially in the currently hot field of AI chatbots. However, for many non-professionals, the contents of Transformer are
2024-08-11
comment 0
929
NetEase Fuxi won the CVPR 2023 UG2+ and VizWiz competitions, and his paper was selected as TIP
Article Introduction:Recently, the results of the CVPR2023 competition were announced. NetEase Fuxi Lab achieved first place in the UG2+ Haze Target Recognition Challenge and VizWiz Few-Sample Target Recognition Challenge of CVPR2023. Their related papers have also been accepted by TIP, the top international journal. This shows that NetEase Fuxi's top technological innovation capabilities in the field of computer vision have been highly recognized internationally. From February to June 2023, the IEEE/CVF Computer Vision and Pattern Recognition Conference (CVPR), as the top conference in the field of international computer vision and pattern recognition, cooperates with authoritative academic institutions and well-known enterprises around the world to hold
2024-01-23
comment 0
846
TAL launches self-developed large-scale mathematics model MathGPT to achieve personalized teaching through AI
Article Introduction:On August 24, 2021, Tian Mi, TAL's Chief Technology Officer, announced that MathGPT, a 100-billion-level large-scale model in the field of mathematics independently developed by TAL, has been officially launched and has begun public testing. Starting today, users can apply to register an account through the official website, try out and experience the model for free. In May this year, TAL announced that it was developing a large self-developed mathematical model, which was named MathGPT. According to reports, MathGPT is a large-scale model for mathematics enthusiasts and scientific research institutions around the world. Its core is problem-solving and problem-telling algorithms, focusing on the vertical field of mathematics. This is also the first large-scale model in China specially built for the field of mathematics. When using MathGPT, users only need to upload math questions through text or pictures to get
2023-08-25
comment 0
1656
Deep learning heart sound classification based on logarithmic spectrogram
Article Introduction:This paper is very interesting. It proposes two heart rate sound classification models based on the logarithmic spectrogram of the heart sound signal. We all know that spectrograms are widely used in speech recognition. This paper processes the heart sound signal as a speech signal and achieves good results. It divides the heart sound signal into frames of consistent length and extracts its logarithmic spectrogram features. The paper proposes long short-term memory (LSTM) and convolutional neural network. (CNN) Two deep learning models classify heartbeat sounds based on extracted features. Imaging diagnosis of the heart sound data set includes cardiac magnetic resonance imaging (MRI), CT scan, and myocardial perfusion imaging. The disadvantages of these technologies are also obvious: high requirements for modern machinery and professionals, and long diagnosis time. The data set used in the paper is a public data set, which contains
2023-09-29
comment 0
1360
AI weaponization becomes a hot topic on underground forums
Article Introduction:According to conventional wisdom, a drive-driven attack is defined as the automatic download of malicious files from a compromised website without user interaction. However, in the majority of cases reviewed during the reporting period, user action was involved - facilitating initial access in more than 30% of incidents. Threat actors use AI to automate attacks The use of artificial intelligence to accelerate these attacks is receiving increasing attention on major cybercrime forums, and interest in weaponizing the technology is growing. Researchers discovered criminal alternatives to mainstream chatbots, such as FraudGPT and WormGPT, in the specialized AI and machine learning sections of these websites, and suggested the use of these options to develop simple malware and distributed denial-of-service (DDoS) queries. AI systems are now available
2024-03-29
comment 0
1119
Baidu Wenxinyiyan launches independent App: only Android version is available for now
Article Introduction:The latest news is that Baidu’s chatbot Wenxinyiyan has been in beta testing for more than a month. Netizens discovered that Wen Xinyiyan has launched a dedicated independent app for internal testing, which supports voice input. Currently, it is only available in the Android version. Wenxin Yiyan (English name: ERNIEBot) is Baidu's new generation of knowledge-enhanced large language model and a new member of the Wenxin large model family. It can interact with people, answer questions, assist in creation, and help people obtain information and knowledge efficiently and conveniently. and inspiration. Wenxinyiyan is a knowledge-enhanced large language model. Based on the Feipiao deep learning platform and the Wenxin knowledge-enhanced large-scale model, Wenxinyiyan continues to integrate learning from massive data and large-scale knowledge, and has the technical characteristics of knowledge enhancement, retrieval enhancement, and dialogue enhancement. For more information, please pay attention to this site.
2024-03-04
comment 0
801
How to install win7 operating system on computer
Article Introduction:Among computer operating systems, WIN7 system is a very classic computer operating system, so how to install win7 system? The editor below will introduce in detail how to install the win7 system on your computer. 1. First download the Xiaoyu system and reinstall the system software on your desktop computer. 2. Select the win7 system and click "Install this system". 3. Then start downloading the image of the win7 system. 4. After downloading, deploy the environment, and then click Restart Now after completion. 5. After restarting the computer, the Windows Manager page will appear. We choose the second one. 6. Return to the computer pe interface to continue the installation. 7. After completion, restart the computer. 8. Finally come to the desktop and the system installation is completed. One-click installation of win7 system
2023-07-16
comment 0
1168
php-insertion sort
Article Introduction::This article mainly introduces php-insertion sort. Students who are interested in PHP tutorials can refer to it.
2016-08-08
comment 0
1024
Graphical method to find the path of the PHP configuration file php.ini, _PHP tutorial
Article Introduction:Illustrated method to find the path of the PHP configuration file php.ini. Graphical method to find the path of the PHP configuration file php.ini. Recently, some bloggers asked in which directory php.ini exists? Or why does it not take effect after modifying php.ini? Based on the above two questions,
2016-07-13
comment 0
773
Huawei launches two new commercial AI large model storage products, supporting 12 million IOPS performance
Article Introduction:IT House reported on July 14 that Huawei recently released new commercial AI storage products "OceanStorA310 deep learning data lake storage" and "FusionCubeA3000 training/pushing hyper-converged all-in-one machine". Officials said that "these two products can train basic AI models." , industry model training, and segmented scenario model training and inference provide new momentum." ▲ Picture source Huawei IT Home compiled and summarized: OceanStorA310 deep learning data lake storage is mainly oriented to basic/industry large model data lake scenarios to achieve data regression from Massive data management in the entire AI process from collection and preprocessing to model training and inference application. Officially stated that OceanStorA310 single frame 5U supports the industry’s highest 400GB/s
2023-07-16
comment 0
1504