Zeyu Liu is a researcher at the Hong Kong Polytechnic University, currently collaborating with Professor Hongxia Yang on multimodal large language models (MLLMs). His recent work focuses on developing MLLMs with strong reasoning capabilities at the 3B and 7B parameter scales. Zeyu began his master’s degree at Harbin Institute of Technology (Shenzhen) in 2022, under the supervision of Professor Qing Liao, specializing in machine learning, particularly multi-task learning and curriculum learning. During his graduate studies, he concentrated on optimization techniques for multi-task learning and developed joint scheduling frameworks that integrate task allocation with data optimization. He authored four research papers, including two as the first author. Building on this foundation, Zeyu has since expanded his research into the domain of large-scale multimodal models, contributing to advancements in reasoning-enhanced artificial intelligence.