Jianxiang He / 何建翔

MPhil in Artificial Intelligence at HKUST(GZ), currently working on multimodal large language models, long video understanding, and efficient key-frame selection.

Multimodal LLM Video Understanding Reinforcement Learning NLP

Email: jhe307@connect.hkust-gz.edu.cn

Affiliation: AI+ Lab / HKUST-GZ NLP Group

I am actively seeking a PhD position for Fall 2026. Feel free to reach out for collaboration or opportunities.

About

I am a Master of Philosophy student at the Hong Kong University of Science and Technology (Guangzhou), advised by Professor Hui Xiong and Assistant Professor Xuming Hu. My current research focuses on long-video understanding, where I explore key-frame selection and multimodal fusion strategies to improve both efficiency and reasoning quality.

Education

Sept 2020 – June 2024

Shandong University

B.E. in Automation · Core courses in NLP, image processing, and machine learning.

Sept 2024 – Present

Hong Kong University of Science and Technology (Guangzhou)

MPhil in AI · Focused on Multimodal LLMs, NLP, and reinforcement learning.

Research Interests

Multimodal LLM

Building unified systems that align language, vision, and audio for robust real-world understanding.

Reinforcement Learning

Designing adaptive optimization strategies for sequential decision-making and efficient policy learning.

Video Understanding

Studying temporal reasoning and sparse key-frame retrieval in long videos for scalable model inference.

Applied NLP

Exploring efficient and reliable language technologies for research, education, and productivity tools.

News & Publications

ACL 2025 A Survey of Mathematical Reasoning in the Era of Multimodal Large Language Model: Benchmark, Method & Challenges

Yibo Yan, Jiamin Su, Jianxiang He, et al.

IEEE ROBIO 2022 The development of spiking neural network: A review

Jianxiang He, Yanzi Li, Yingtian Liu, et al.

Visitor Locations