Reporter:Hong Kong University of Science and TechnologyProfessor Chen Kai
Reporting location:Messagefloor114Conference Room
Report time:2024Year8month24Day (weeksix）下注noon3Point
Report title:Towards AI-centric Networking: Challenges and Opportunities
Personal introduction: Chen Kai, Professor, Department of Computer Science and Engineering, Hong Kong University of Science and Technology，Intelligent Network and System Laboratory(iSING Lab)Director, HKUST-Director of Inspur Advanced Cloud Intelligent Technology Joint Laboratory, Hong Kong University of Science and Technology-Director of WeChat Artificial Intelligence Technology Joint Laboratory, Hong Kong University of Science and Technology-Deputy Director of China Unicom Joint Laboratory, Hong KongRGCChief Scientist of Theme Research Program。Mainly engaged in data center networking、High performance network、AIResearch on computing power networks, machine learning systems, etc., on many occasionsACM SIGCOMM, USENIX NSDI/OSDI, IEEE/ACM TONPublished articles in top international academic conferences and journals, GetIEEE ICNP 2023Best Paper Award andACM SIGCOMM 2010Best Paper Nomination Award, served asSIGCOMM, NSDI, EuroSys, TONProgram member and seabet sports betting editorial board member of other international conferences and journals，is an Asia Pacific webinarAPNetInitiator and Steering Committee Chairman.
Abstract:The ever-growing AI and ML workloads present unprecedented challenges as well as opportunities for designing AI-centric networking for AI clusters. In this presentation, I will first talk about the special characteristics of communication with the distributed AI/ML training and then discuss how to explore these characteristics in designing next-generation network architecture and protocols for AI and ML workloads.
Reporter:Researcher Chen Kai, Chinese Academy of Sciences
Reporting location:Messagefloor114Conference Room
Report time:2024Year8month24Day (weeksix）下注noon4Point
Report title:Large Model Security: Opportunities and Challenges
Personal introduction: Chen Kai，National leading talents in science and technology innovation，Researcher at the Institute seabet online sports betting of Information Technology, Chinese Academy of Sciences、Doctoral Supervisor，Director of Frontier Innovation and Science and Education Integration Center，Director of the Editorial Department of "Journal of Information Security"。Won the Young Scientist Award of the Chinese Academy of Sciences、CCF-IEEE CSYoung Scientist Award、Beijing Science and Technology Award for Outstanding Youth Zhongguancun Award、NASACYouth Software Innovation Award、Beijing “Outstanding Youth” Fund、Beijing Zhiyuan Young Scientist and many other honors。Main research areas include system security、Artificial Intelligence Security。atIEEE S&P、USENIX Security、ACM CCS、NDSS、ICSEPublished academic papers at high-level conferences and journals seabet app download in the field of information security200Multiple articles。Host the major research project of National Natural Science Foundation of China、Key projects, etc.。
Abstract:The emergence of large models has had an important impact on security research。On the one hand, it can empower traditional offensive and defensive methods，Make it smarter and more efficient；On the other hand，Its own security issues，such as"Jailbreak" problem、"Illusion" problem、Privacy issues, etc.，Also raises concerns when using large models。This report will focus on these two aspects，Focus on how to use large models to empower traditional security offense and defense，We will also talk about the safety issues of large models themselves。
