I am a first-year PhD student at City University of Hong Kong, and I am very fortunate to be advised by Prof. Kaidi Xu. I completed my master’s study at Wuhan University, where I was advised by Prof. Chong Teng.
My research interests focus on uncertainty estimation for language models and improving reliability in multi-step decision-making tasks. In particular, I study how language models can recognize and manage uncertainty across MoE routing, tool use, and agentic planning, where models need to make a sequence of decisions rather than a single prediction.
Academic Awards
Our work DiscoSG received the EMNLP 2025 Outstanding Paper Award. I am especially grateful to Prof. Zhuang Li for the collaboration and his valuable support on this work.
Collaboration
I warmly welcome research collaborations in related areas. I am also happy to work with junior students who are interested in starting research on LLMs. Please feel free to reach out at kyrielincn@gmail.com.
Publications
(EMNLP 2025, Outstanding Paper Award, Oral Presentation) Shaoqing Lin, Chong Teng†, Fei Li, Donghong Ji, Lizhen Qu, Zhuang Li. “DiscoSG: Towards Discourse-Level Text Scene Graph Parsing through Iterative Graph Refinement.” Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, November 2025. [Code] [Paper]
(Under Review) Yongqi Tong, Ruirui Wang, Kewei Fu, Shaoqing Lin, Sijie Dong, Zhenyu Zhang, Jiang-Ming Yang, Xin Zhang, Jianshe Li. “MOSAIC: Scaling Multi-Objective Alignment via Stability-Gated Reverse Curriculum and Adaptive Goal Expansion.”
Internships
Ant International, Research Intern, Dec 2025 - Present.
Microsoft, Research Intern, Apr 2025 - Sep 2025.
Xiaomi, Algorithm Engineer Intern, Jan 2025 - Apr 2025.
Education
- City University of Hong Kong, 2026 - 2030 (expected). First-year PhD student. Advisor: Prof. Kaidi Xu.
- Wuhan University, 2023 - 2026. M.S., Cyber Security. Advisor: Prof. Chong Teng.
- Shandong University, 2019 - 2023. B.S., Computer Science and Technology.