Donghang Duan | 段东杭

I am currently an undergraduate student at the Yingcai Honors College of the University of Electronic Science and Technology of China, advised by Prof. Xu Zheng. I am also an incoming Ph.D. student at the College of Computer Science and Artificial Intelligence, Fudan University, in a joint training program with the Shanghai Artificial Intelligence Laboratory under the mentorship of Dr. Dongrui Liu.

Google Scholar / GitHub / Email / Zhihu

Research

My research interests lie in trustworthy AI, with a focus on LLM safety and alignment, agent security, and privacy. I think it is one of the most important challenges to make AI both useful and reliable in the future. My previous work spans robust and blockchain-based federated learning.

	FedWeave: Rethinking the Unit of Specialization in Heterogeneous Federated MoE-LoRA Donghang Duan, Xu Zheng, Lizong Zhang, Chong Mu, Meng Han arXiv preprint, 2026 paper / code FedWeave combines prototype-level expert aggregation with client-level router optimization to handle heterogeneous tasks in federated MoE-LoRA.
	Look Twice before You Leap: A Rational Framework for Localized Adversarial Text Anonymization Donghang Duan, Xu Zheng, Yuefeng He, Chong Mu, Leyi Cai, Lizong Zhang Findings of ACL, 2026 paper / code A localized, training-free framework balances privacy gain and text utility through an attacker, arbitrator, and anonymizer.
	Exploiting Parasitic Dependency for Free-Rider Elimination in Blockchain-Based Federated Learning Donghang Duan, Xu Zheng, Yifu Zheng, Chong Mu, Ruozhou Wang, Ke Yan IEEE IPCCC, 2025 paper An incentive and reputation-based framework counters advanced free-rider attacks in blockchain-based federated learning.

Academic Service

[Reviewer] ICLR 2027

Last updated: 2026.9.26