Yang Penghui

Yang Penghui

Tsinghua University.

About Me

👋 Hi there!

I’m Penghui Yang (杨鹏晖), a senior undergraduate student majoring in Building Environment and Energy Engineering at Tsinghua University.

I will begin my Ph.D. journey in 2026 at the Shanghai Artificial Intelligence Laboratory (Shanghai AI Lab), under the supervision of Yuhang Zang, in a joint training program with the School of Artificial Intelligence, Shanghai Jiao Tong University (SAI).

My research interests broadly lie in:

  • Deep Learning
  • Multimodal Learning
  • Large Language Models (LLMs)

Currently, I’m particularly interested in vision-language pre-training and post-training alignment of large multimodal models.

📄 For more details, please see my CV
📬 Email • GitHub

📰 News

  • 2025.10.21 — Started my internship at Shanghai AI Lab (Beijing Base).
  • 2025.10.10 — Honored to receive the National Scholarship.
  • 2025.05 — Won the Second Prize in the Tsinghua University Artificial General Intelligence Competition for a RAG project.

🏆 Honors & Awards

  • National Scholarship, Ministry of Education of China (2025.10)
  • Second Prize, Tsinghua University Artificial General Intelligence Competition (2025.05)
  • Comprehensive Excellence Scholarship (Top 10%), Tsinghua University (2024.12)
  • Comprehensive Excellence Scholarship (Top 10%), Tsinghua University (2023.12)

🎓 Education & Experience

B.Eng. (2022.08 – 2026.06, expected)
Department of Building Environment and Energy Engineering, Tsinghua University

  • Focused on AI applications in building environment and energy systems.
  • Rank: 1/16 in my major.

Intern (2025.10 – Present)
Shanghai Artificial Intelligence Laboratory (Beijing Base)

  • Exploring multimodal large language models (MLLMs), with a focus on vision-language pre-training and post-training alignment.

Ph.D. Student (2026.09, expected)
Shanghai AI Lab & School of Artificial Intelligence, Shanghai Jiao Tong University (Joint Program)

💡 Research Interests

I am broadly interested in foundation models and multimodal intelligence, especially topics at the intersection of representation learning, model efficiency, and alignment.
My long-term goal is to build trustworthy and efficient general-purpose AI systems that can adapt across modalities and domains.

🌱 Outside of Research

When I’m not coding or writing, I enjoy:

  • 🏃 Fitness, basketball, and billiards — keeping balance between body and mind.
  • ☕ Café hopping — good ideas often come with good coffee.