Dynamic Cross-Modal Prompt Generation for Multimodal Continual Instruction Tuning
Conference on Neural Information Processing Systems
PaperPersonal academic homepageLAMDA / NJU
Learning continually.
Connecting modalities.
Exploring continual learning, pre-trained model reuse,
and multimodal large language models.
About me
01 /I am an M.Eng. student at the School of Artificial Intelligence, Nanjing University, and a member of the LAMDA Group led by Prof. Zhi-Hua Zhou.
I received my B.Eng. degree from the School of Artificial Intelligence, Nanjing University in June 2025. In the same year, I began pursuing my M.Eng. degree at Nanjing University, under the supervision of Dr. Da-Wei Zhou.
Research interests
02 /My research centers on machine learning and its applications in computer vision, with a focus on:
IContinual
Continual
Learning
IIPre-trained
Pre-trained
Model Reuse
IIIMultimodal Large
Multimodal Large
Language Models
Publications
03 /Conference papers
2026
Hierarchical Semantic Tree Anchoring for CLIP-Based Class-Incremental Learning
Conference on Neural Information Processing Systems
Paper
BOFA: Orthogonal Low-Rank Bridge-Layer Fusion for CLIP-Based Class-Incremental Learning
Proceedings of the AAAI Conference on Artificial Intelligence
PaperPreprints
2026Get in touch
04 /- hut@lamda.nju.edu.cn
- Office
- A321, School of Artificial Intelligence, Xianlin Campus
- Address
- National Key Laboratory for Novel Software Technology, Nanjing University, Xianlin Campus Mailbox 603, 163 Xianlin Avenue, Qixia District, Nanjing 210023, China