avatar

Guanyao Mao

4th Undergraduate
MIT,
Fudan Univerisity
220 Handan Road, Yangpu District, Shanghai, 200433
adiontenasagna@gmail.com
22307130309@m.fudan.edu.cn


Short Bio

I am an undergraduate student at Fudan University, advised by Prof. Hao Zhang. Currently, I work with the Computational Design and Fabrication Group (CDFG) at MIT CSAIL, under the supervision of Prof. Wojciech Matusik. During the summer of 2025, I was a visiting student at Stanford University, where I deepened my knowledge in machine learning and computational modeling.

Research Interest

-AI for Scientific Methodology and Mechanism Inference My research aims to redefine the core paradigm of AI for Science, shifting the focus from Correlation to Mechanism Inference. I am driven to build Autonomous Scientific Discovery Systems that functionally mimic the human scientific mind. This involves integrating the flexibility of Intuition (LLM-driven hypothesis generation) with the rigidity of Formal Theory (mechanistic law refinement). My work tackles the fundamental question of how to enable AI not just to predict outcomes, but to fundamentally understand and correct the underlying governing mechanisms of complex systems.

-Neuro-Symbolic Hybrid Architecture for System Rigor I specialize in designing next-generation Neuro-Symbolic Hybrid Architectures to address the inherent limitations of Large Language Models (LLMs) in scientific logic and constraint handling. The core of this effort is the construction of tightly coupled, highly constrained systems that necessitate Simulation-Grounded Verification. This involves developing language interfaces (e.g., AxiomIR) that bridge the LLM’s fluidity with the strict requirements of physical constraints, and integrating rigorous physics simulators to ensure that AI-generated mechanistic hypotheses are robust in both logic and physics, thereby establishing Trustworthy Scientific AI.

-System Generalization and Optimization in Materials Engineering My research is dedicated to generalizing the principles of autonomous mechanism discovery to complex Engineering System Optimization and Materials Discovery. I focus on how to successfully scale micro-level mechanistic laws (e.g., at the molecular/material level) into macro-level system governing principles. This direction leverages my foundational experience in advanced materials (such as energy systems) and extends into domains like computational catalysis, structural robustness, and mechanistic inference in environmental or biomolecular systems, with the ultimate goal of developing scientific foundation models capable of continuous learning and self-improvement.

Honors and Awards

Know more about Me

Publications
  1. ZiJie Qiu, Luozhijie Jin, Zijian Du, Hongyu Chen, Guanyao Mao, Yan Cen, Siqi Sun, Yongfeng Mei, Hao Zhng
    NPJ Computational Materials, 2024

  2. Guanyao Mao, Zhiyang Dou, Minghao Guo, Benjamin Tod Jones, Wojciech Matusik
    Undereview ICML2026

  3. Zijian Du, Luozhijie Jin, Guanyao Mao, Hongyu Chen, Yan Cen, Yongfeng Mei, Hao Zhang
    Preprint

  4. Projects

    1. Research project
      Guanyao Mao
      This project focuses on bulk tellurium materials with triangular lattice van der Waals molecular chain structures, calculating the electrical and mechanical physical properties of three-dimensional bulk tellurium, and exploring its potential applications in the field of thermoelectric conversion.

    2. Research Project and to be submitted
      Guanyao Mao, Yuqian Zhang, Yun Wang
      This project is based on large language model (LLM) technology and focuses on automated knowledge extraction from open-source perovskite scientific literature and related material property data to build a perovskite intelligent agent.

    3. Products and Patent
      Guanyao Mao, Yuqian Zhang, Ruixiao Lian, Hao Zhang
      This project utilizes Fresnel lenses and magnesium-based thermoelectric materials to achieve all-angle utilization of solar energy.

    4. Contact

      Address:220 Handan Road, Yangpu District, Shanghai, 200433
      Office Location: Genetics Building, Handan Campus, Fudan University
      Email: 22307130309(at)fudan.edu


Last Update: