About Me
Hi, I'm Xiaoyu Fan, an undergraduate student at Huazhong University of Science and Technology, currently in my third year of Intelligent Medical Engineering.
I am currently a research assistant at CAI, THU, advised by Prof. Tianyu Liu. Previously, I was visiting at the DeepDelta Lab, SJTU, advised by Prof. Shaobo Cui; and undertook research training with the DeepKW team at the School of Journalism and Information Communication, HUST, advised by Prof. Ran Wang.
My research centers on Human-AI Synergy, focusing on communication and co-operation amongst multi-agents, controllable (creative) language generation, collaboration between AI & human scientists, and the training & evaluation of foundational models at various stages. Additionally, I'm passionate about exploring large-scale ML driven by a vast amount of high-quality data, and even more curious about uncovering the mechanistic interpretability behind the empirical phenomena: why models train, where behaviors emerge, and when scaling rules break.
Research Interest: Multi-agent Systems, AI4Medicine, and Representation Learning.
Download my CV (Updated: Sep. 2026)
Education
Huazhong University of Science and Technology
School of Artificial Intelligence and Automation
B.Eng. in Intelligent Medical Engineering
2024 - 2028
Research Experience
DeepDelta Lab, Shanghai Jiao Tong University
Visiting Student
Hosts: Prof. Shaobo Cui
Jul. 2026 - Aug. 2026
News
- [09/2026] My application for the "Rising Stars Poster" was accepted. I will attend X-AGI 2026 in October and look forward to learning and exchanging ideas!
- [08/2026] Our CWF paper is accepted to EMNLP 2026 as a Findings paper.
- [05/2026] Received Meritorious Winner in the 2026 Mathematical Contest in Modeling.
- [09/2025] Received the Self-Reliance and Endeavour Scholarship from Huazhong University of Science and Technology for the 2024-2025 academic year.
Publications
CWF: A Collaborative Writing Framework for Personalized and Reliable Popular Science Writing
EMNLP 2026 Findings
Selected Projects
Omission-based Deception Fake News Detection Dataset Apr. 2026
Constructed a benchmark dataset of approximately 4,000 samples for detecting omission-based deception in LLMs, with hierarchical labels and six fine-grained categories. Implemented data generation scripts and evaluation workflows supporting SFT and prompting-based approaches.
Image Segmentation and Region Coloring System Mar. 2026 - Apr. 2026
Developed an end-to-end C++ / OpenCV image processing pipeline for seed generation, watershed segmentation, region adjacency graph construction, and graph coloring, achieving second-level runtime on 600 x 600 images with about 1,000 seeds.
Honors & Awards
- Science and Technology Innovation Scholarship, HUST (University-wide)2025 - 2026
- Silver Award at the "Taikang Cup" Interdisciplinary Innovation Competition in Medicine&Engineering2026
- Meritorious Winner in the 2026 Mathematical Contest in Modeling2026
- Self-Reliance and Endeavour Scholarship, HUST (University-wide)2024 - 2025
Services
- Reviewer @AAAI 2026, @EMNLP 2026, @IJCAI 2026.
Academic Activities
- 2026 X-AGI Conference Rising Stars Poster PresenterOct. 17 - 18, 2026
Misc.
· Hobbies: 🏀、🏐、NBA、Board Game、Science Fiction