Skip to main content Start main content

Department of Computing

Research Assistant

(Ref. 260703009)

[Appointment period: twelve months]

Duties 

The appointee will assist the project leader in the research project – “Collaborative Generative AI”.  He/She will be required to:

(a)    investigate cutting-edge techniques for foundation large language models and multimodal large language models, with emphasis on pre-training, post-training and model merging/fusion;

(b)    develop secure, scalable and efficient infrastructure for data management, distributed training, model evaluation, deployment and ongoing model improvement;

(c)    construct and maintain large-scale distributed training systems for large language models, including cluster management, fault-tolerant scheduling, mixed-precision training and end-to-end pipeline optimisation to enable efficient pre-training and fine-tuning at scale;

(d)    explore efficient architectural designs for both dense and Mixture-of-Experts (MoE) models, including but not limited to MLA, mHC, Loop Transformer and other parameter-efficient or compute-efficient structures, aimed at enhancing model capacity and training efficiency under constrained computational budgets; and

(e)    perform any other duties as assigned by the project leader, the Head of Unit or their delegates.

Qualifications

Applicants should have:

(a)    an honors degree or an equivalent qualification, preferably in Computer Science, Artificial Intelligence, Data Science Engineering or a related discipline;

(b)    demonstrated expertise in foundation models, generative AI, multimodal learning, machine learning systems, model merging/fusion, or closely related fields; 

(c)    strong capacity for self-directed learning; and 

(d)    commitment to open, reproducible and ethically responsible AI research.

Applicants are invited to contact Prof. Hongxia Yang at telephone number 2766 7247 or via email at hongxia.yang@polyu.edu.hk or Mr Jianmin Wu via email at jianmin.wu@polyu.edu.hk for further information.

Conditions of Service

A highly competitive remuneration package will be offered. 


The closing date for application is 10 July 2026.



Posting date: 3 July 2026