Yuedong Xu
Affiliations: College of Computer Science and Artificial Intelligence, and Artificial Intelligence Innovation & Incubation Institute, Fudan University (复旦大学计算与智能创新学院/人工智能创新与产业研究院).
-
Yuedong Xu 徐跃东
-
Professor, Advisor of MediaNet Group
-
Office: Interdisciplinary Science Building, Jiangwan Campus, Fudan University
-
E-mail: ydxu@fudan.edu.cn
-
Research Interests:
- LLM training and inference systems
- Parallel pretraining systems (并行化机制、超节点系统)
- Reinforcement learning systems (强化学习后训练系统加速、多模态强化学习系统)
- Inference and Agentic systems (推理与智能体系统优化)
- Multi-modal LLM alignment (多模态安全对齐)
Yuedong Xu is a Professor with College of Computer Science and Artificial Intelligence, Fudan University, China. He received the BS degree from Anhui University, the MS degree from Huazhong University of Science and Technology, and the PhD degree from The Chinese University of Hong Kong. After graduation, he worked as a MENRT Postdoctoral fellow in INRIA Sophia Antipolis (not far away from Nice and Cannes) and Université d’Avignon, France (a small but beautiful city famous for its Lavender and Festival d’Avignon). He received the OKAWA Foundation research grant in 2019, and several teaching awards such as Shanghai Tang-Junyuan educational foundation teacher award (2023), Fudan BYD teacher award (2024), Shanghai Program of Shanghai Academic/Technology Research Leader (2025). He serves as an associate editor for IEEE Transactions on Network Science and Engineering. His areas of interests include foundation model training and inference systems, and multimedia networking. He has published a number of papers in premier conferences and journals including USENIX NSDI, USENIX ATC, ACM Mobisys, ACM CoNEXT, ACM Mobihoc, IEEE Infocom, IEEE/ACM ToN and IEEE JSAC.
欢迎本科生、硕士生、博士生,以及国家人工智能学院(深圳河套学院、上海创智学院等)联合培养博士生加入课题组!我们的研究和工业界的实际需求和前沿技术结合密切,科研实习机会比较多(例如阿里、蚂蚁、腾讯、微软、华为等)
Q1: 招生名额及方向?
复旦大学学术学位博士生(AI Infra方向)、集成电路专项推免硕士生(AI Infra方向)、复旦大学 + 国家人工智能学院联合培养博士生(AI Infra/Trustworthy AI方向皆可)
Q2: 往届学生去向?
博士生:暂无更多信息,因为2023年才有独立招收的博士生入学。合作指导博士生比较努力和优秀,获得ACM Sigcomm China优秀博士论文、上海市信息学会优秀博士论文,毕业后在海外从事博士后研究或以优才方式加入央企研究院。硕士生:近5年以阿里、字节、腾讯、华为、选调生、美国读博为主,也有少数同学拒掉头部互联网企业Offer令觅他处,不少人获得复旦大学优秀毕业研究生、上海市优秀毕业研究生、国家奖学金。(仅仅代表过去状况)
Q3: 招生要求?
自驱力强,认真踏实,阳光乐观,能听取不同意见,不眼高手低,不妄自菲薄,有毅力能够完整地做完一项工作,具有团队合作精神。AI Infra能力要求:熟悉1门编程语言(C++/Python/Go等之一)、编程能力较强,有MLSys的项目经验是个加分项。可信AI:基础数学课成绩不错、熟悉Python编程,或者作为主力参加过重要的竞赛或相关科创项目。
Q4: AI Infra没学过,自学是否困难?
需要一定的基础,但没有想象中的困难。目前本人刚转入计算与智能创新学院/AI3院,课题组尚无一人是计算机本科出身。参考课程:加州伯克利的《CS294-162 Machine Learning Systems》、卡内基梅隆的《15-442/15-642: Machine Learning Systems》、斯坦福的《CS 329S: Machine Learning Systems Design》、深圳河套学院的《Machine Learning System》https://mlsys-course.github.io、上海创智学院的《Systems for Artificial Intelligence》https://syfeng.net/pages/teaching.html 。此外,复旦大学也有华为的实训基地和实训课程。
news
| Jan 24, 2026 | One paper on multi-modal LLM alignment was accepted to ICLR 2026. |
|---|---|
| Dec 20, 2025 | One collaborative paper was accepted to ToN. |
| Dec 09, 2025 | Our work on checkpointing and failure recovery for LLM training is accepted to IEEE Infocom 2026 |
| Dec 07, 2025 | One collaborative paper was accepted to NDSS 2026. |
| Sep 19, 2025 | Our work on Federated LLM Finetuning is accepted to NeurIPS 2025 |