Jianqiao Zeng

I am a PhD student at AI³ Institute, Fudan University and Shanghai Academy of AI for Science (starting 2026.9), jointly advised by Prof. Hao Xiong (Fudan) and Prof. Junchi Yan (SJTU).

My research centers on deep learning architectures, representation learning, and AI agents — building models and systems that reason over structured signals and decompose complex problems.

Research

I work on three interconnected directions:

• Model Architectures. Designing efficient and expressive neural architectures — including co-evolutionary state–mesh updates (CoEvol-NO) and frequency-domain position encoding (FreqONet).
• Representation Learning. Studying how neural networks encode functions, frequencies, and structure across modalities; theoretical analysis of approximation and stability.
• AI Agents. Building structured-reasoning agents that decompose complex problems into tractable sub-tasks (e.g., the Learning Agent system).

Education

• Incoming Ph.D. in Computer Science, AI³ Institute, Fudan University, 2026.9 – present
• B.E. in Artificial Intelligence, Sun Yat-sen University (中山大学), 2022 – 2026