Xungang Gu is currently a Ph.D. student at Deakin University and works as a Natural Language Processing (NLP) algorithm engineer. He specializes in artificial intelligence and NLP, with a research focus on Human-Centered Large Language Models (LLMs) and adaptive training techniques for large-scale models. His work bridges academic research and industrial applications, contributing to the development of more aligned and efficient LLM systems. Prior to his doctoral studies, he accumulated experience in applied NLP research and development, particularly in model adaptation and instruction tuning. His academic interests lie in enhancing model usability, robustness, and human alignment in AI systems.
ZHANG He, Associate Researcher in Computer and Information Technology, Chief Scientist at Kexin Technology. Member of CSIG Technical Committee on Document Image Analysis and Recognition, and Senior Member of the China Computer Federation. Research expertise: Natural Language Processing, Computational Simulation. He has participated in five key national R&D plans and major special projects, one National Natural Science Foundation of China project, and was responsible for implementing one "Top Talent" project from Beijing Municipal Science and Technology Commission. He has published 26 academic papers and applied for 22 invention patents. He has served as Keynote Speaker or Roundtable discussion guest at important international conferences such as WAIC and CPT.
Ruohua Xu is the General Manager at Kexin Technology. His expertise spans business management and natural language processing. He is particularly focused on the application of large language models, including Retrieval-Augmented Generation (RAG) and AI agents, promoting the integration of advanced NLP technologies in practical business solutions.
Dr. Ming Liu is a Senior Lecturer at the School of
Information Technology, Deakin University, Australia. He holds a Ph.D. from the
Faculty of IT, Monash University. He is a member of the Association for
Computational Linguistics. He proposed the “learn to actively learn” approach
for active learning and developed a few efficient text summarization
models/pipelines (e.g., SummPip, SciSummPip, GLIMMER), most of which are widely
used in low-resource text generation settings. He has a strong interest in solving
real-world text mining problems, particularly in domain-specific settings.
