AI Research Engineer · Agentic AI & Reliable MLMaster of Philosophy (Engineering) · The University of Sydney · DUAL Group
Hi, I’m Zerun. I’m an MPhil researcher at the University of Sydney, working on machine learning when data is limited, devices have different resources, and communication is unreliable. My work spans algorithm design, implementation, and experiments.
My research looks at how machine learning behaves with limited labels, unreliable wireless connections, and devices with different capabilities. Here are some of the projects I’ve worked on.
How can we preserve the information an AI model needs when a wireless connection is noisy or bandwidth is limited? WaSeCom studies this under changing channel conditions.
Contributed modelling, experiment design, robust training simulations, and evaluation under constrained channels.
IEEE Journal on Selected Areas in Communications · DOI ↗
FeDEQ explores how devices can train a model together when they have different data, memory, energy, and communication limits, using deep equilibrium models.
Implemented FeDEQ components, explored communication-efficient aggregation, and evaluated non-IID NLP and vision settings.
IEEE Internet of Things Journal · under review
With only a few annotations, a model can become more confident without gaining new evidence. BRAVE controls how earlier estimates feed into later updates to reduce this effect.
Zerun Niu — first author; led algorithm design, literature review, experimental design, code implementation, and experimental deployment.
Transactions on Machine Learning Research · under review
A website for the DUAL Group’s people, publications, and projects, designed to make the group’s work easy to find and keep up to date.
Designed, developed, deployed, and maintain the group's official web presence and research communication system.
Websites and tools I build and maintain alongside my research.
I built this site to bring my research, projects, and experience together. It runs on Astro, with automated checks and deployment to help me keep it up to date.
Source ↗An AI guide to my research and experience. You can type a question or use voice to explore the work on this site.
Meet Digital Zerun ↓Designed, built, and maintained end to end for the DUAL Group's public research presence at the University of Sydney.
Live site ↗Ongoing research and teaching across the University of Sydney, UNSW Sydney, and the DUAL Group.
education · The University of Sydney · Research in reliable and efficient AI systems, supervised by Prof. Nguyen Tran in the DUAL Group; nearing completion.
teaching · UNSW Sydney · I help students understand computing concepts, work through practical problems, and learn from feedback.
teaching · The University of Sydney · I lead tutorials and help students work through questions, discuss ideas, and solve problems together.
research · DUAL Group, The University of Sydney · Research engineering and experimentation for distributed learning and efficient AI systems.
education · The University of Sydney · Graduated with Distinction; final-year work centred on machine learning and distributed systems.
Hi, I’m Digital Zerun, an AI guide to Zerun’s research, projects, and experience. You can type a question or start a voice session. My voice is an authorised AI clone of Zerun’s. I use information he has made public, and I can’t make commitments on his behalf.