Portfolio / Spatial Intelligence
QINLEI
XIE
I aim to build physics-grounded causal world models that can truly understand
the world and guide the development of robotics and
spatial intelligence.
Causal World Models
Physics-Grounded AI
Embodied AI
Robotics
Spatial Intelligence
Current / Status
Undergraduate at
Dalian University of Technology
Big Data Management · Class of 2028
Dalian University of Technology
Big Data Management · Class of 2028
About / Background
ABOUT ME
Hello! I am Qinlei Xie (谢钦磊), an undergraduate at
Dalian University of Technology, majoring in Big Data Management
and Applications.
I previously collaborated with Prof. Yiming Li at
Tsinghua University and Prof. Weiwen Liu at
Shanghai Jiao Tong University. I am currently collaborating with
Zhiwei Yu.
Education
Dalian University of Technology
B.E. in Big Data Management and Applications
2024 – 2028 (Expected)
Experience / Collaborations
RESEARCH
EXPERIENCE
Zhiwei Yu
Collaborating with Zhiwei Yu
Shanghai Jiao Tong University
Collaborated with Prof. Weiwen Liu
Personalization and deep research agents
Research / Focus Areas
RESEARCH
INTERESTS
My research focuses on building physics-grounded causal world models
for AI. I believe these models can truly understand the world and guide the
development of robotics and spatial intelligence.
Causal World Models
Building models that connect perception, action, and cause-effect structure in the physical world.
Physics-Grounded AI
Grounding reasoning in physical constraints so AI can predict, plan, and act with real-world consistency.
Robotics & Spatial Intelligence
Using world understanding to guide robots and spatially intelligent agents in complex environments.
Latest News
2026.06 — Causal Plan is online, moving embodied planning from token prediction toward physically grounded causal reasoning.
2026.03 — GameVerse was accepted to ICML 2026.
2026.03 — GameVerse was accepted to ICML 2026.
Publications / Selected Work
PUBLICATIONS
ARXIV 2026
Token Predictors Are Not Planners: Building Physically Grounded Causal Reasoners
Zheng Lu*, Mingqi Gao*, Qinlei Xie*, Wanqi Zhong, Hanwen Cui, Heng Cao, Zirui Song, Yifan Yang, Chong Luo, Bei Liu†, Yiming Li†
arXiv preprint, 2026 · Equal contribution
Project Highlights
HIGHLIGHTS
PROJECT
Token Predictors Are Not Planners: Building Physically Grounded Causal Reasoners
A physically grounded causal reasoning framework for embodied planning, introducing
Causal-Plan-Bench and Causal-Plan-1M.
PROJECT
GameVerse: Can Vision-Language Models Learn from Video-based Reflection?
An ICML 2026 benchmark for evaluating Vision-Language Models in video game environments,
with a reflect-and-retry paradigm for learning from gameplay failures.
Connect / Get In Touch
LET'S
CONNECT
I'm always open to discussing research ideas, potential collaborations,
and PhD opportunities. Feel free to reach out!
Life / Beyond Research
OUTSIDE
THE LAB
🎹
Piano — Grade 8
Trained in classical piano since childhood. Music provides patience and discipline that carries into research.
🏀
Basketball — Swing Forward
A versatile wing who thrives on sharp-shooting and aggressive drives. Precision and tenacity, on court and in research.
© 2025 Qinlei Xie · Built with spatial curiosity
Drag to rotate · Arrow keys to navigate
