Yangcheng Yu
Undergraduate Student
Department of Electronic Engineering
Tsinghua University
Email | Github

👋 About Me

I am an undergraduate student at the Department of Electronic Engineering, Tsinghua University (2023 - 2027). My research interests lie at the intersection of Embodied AI, World Models for Robotics, and Vision-Language-Action (VLA) Models.

I am currently focused on developing robust world model architectures for embodied planning and long-horizon robotic tasks.

🔬 Selected Research

• MoWM: Mixture-of-World-Models for Embodied Planning (Co-first Author)

  Proposed a hybrid world model architecture that fusions latent motion-aware representations with fine-grained pixel features via latent-to-pixel modulation. This approach addresses limitations of pixel-only models and significantly improves precise action decoding. (ICML 2026 Submission).

• WAM - Action VAE

  Developed a unified Reasoner-Dreamer-Actor architecture to synchronize language, video generation, and robot control. I led the development of the Action VAE model, processing over 10 million trajectories to create a universal action representation across 5+ heterogeneous robotic platforms.

• PAM: A Pose–Appearance–Motion Engine

  Co-designed a decoupled three-stage framework for Sim-to-Real HOI video generation. By augmenting training with PAM-generated synthetic videos, models trained on only 50% real data can match the performance of a full 100% real-data baseline. (CVPR 2026 Submission).

• Impromptu VLA (NeurIPS 2025 Datasets and Benchmarks)

  Curated and distilled a large-scale multimodal driving dataset (80K+ clips) concentrated on unstructured and long-tail urban scenarios.

🎓 Education

• Tsinghua University, Beijing, China B.E. in Electronic Engineering (GPA: 3.75/4.0, Rank: Top 20%).

🏆 Honors and Awards

• Innovation Excellence Scholarship (National Endeavor Scholarship), Tsinghua University 2024, 2025
• 1st Prize, National Undergraduate Physics Competition (Regional Level) 2024
• 2nd Prize, Tsinghua AI Challenge (Team Leader) 2024
• 2nd Prize, Tsinghua Hardware Design Competition (Team Leader) 2024

💻 Internships

Singapore
2026.01 – Present

• Research Intern, S-Lab, Nanyang Technological University

Beijing, China
2025.06 – 2026.01

• Research Intern, Manifold AI

Beijing, China
2025.06 – 2025.09

• Research Intern, Institute for AI Industry Research (AIR), Tsinghua University

Beijing, China
2024.11 – 2025.05

• Research Assistant, Future Intelligence Lab (FIBLab), Tsinghua University

🛠️ Skills

• Programming: Python, C/C++, C#, R, MATLAB, PyTorch.
• Tools: Linux, Git, CUDA, ROS2, Xiaomi Cyberdog platform.