About Me

I am a senior undergraduate majoring in Software Engineering at Nankai University and an incoming PhD student at EPIC Lab, Peking University, advised by Prof. He Wang.

Currently, I work on humanoid robotics as a research intern at GALBOT, advised by Prof. He Wang and Prof. Li Yi. Previously, I was a research intern at THUNLP, Tsinghua University, where I studied multimodal reasoning with Prof. Peng Li and Prof. Yang Liu.

At GALBOT, I explore humanoid robot learning across S2 reasoning, S1 VLA/WAM policies, and S0 whole-body control. My work includes developing humanoid VLA/WAM policies for loco-manipulation, investigating the training of AstraBrain-WBC as a whole-body control foundation model and low-level cerebellum for higher-level policies, and exploring human egocentric data for humanoid learning.

My research focuses on building generalizable embodied agents that can perceive, reason, and act in complex real-world environments.

Please feel free to contact me if you are interested in my work or would like to chat about any topic.

Research Interests

• Humanoid loco-manipulation: coordinated learning and integration of S2 reasoning, S1 VLA/WAM policies, and S0 whole-body control, connecting high-level reasoning with task execution and reliable physical control.

News

• [2026-10] Released PhysRSI-eval, studying physical recursive self-improvement and observational scaling of LLM policies.
• [2026-09] Awarded the National Scholarship.
• [2026-09] Released our technical report on GPT-6 Astra as Embodied Policies.
• [2026-08] EscapeCraft-4D is accepted to EMNLP 2026.
• [2026-06] HumanTracker is accepted to ECCV 2026.
• [2026-05] LIMMT is accepted to ICML 2026.
• [2026-02] Humanoid-GPT is accepted to CVPR 2026.
• [2025-12] Started my research internship at GALBOT.
• [2025-05] Released our work on Visual Abstract Thinking.
• [2024-09] Awarded the National Scholarship.

Publications & Preprints

* Equal contribution; † Corresponding author.

Tech Report

Systematically Exploring the Capabilities of GPT-6 Astra as Embodied Policies

Galbot Team (including Dairu Liu)

Technical report, arXiv, 2026.

Paper

ECCV

HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark

Dairu Liu*, Zekun Qi*, Jiayu Zeng*, Ruixi Yu, Yu Guan, Yintianrun Zhang, Xuchuan Chen, Sikai Liang, Zekai Li, Chenghuai Lin, Xinqiang Yu, Wenyao Zhang, He Wang†, Li Yi†

European Conference on Computer Vision (ECCV), 2026.

Paper Project Code

CVPR

Humanoid-GPT: Scaling Data and Structure for Zero-Shot Motion Tracking

Zekun Qi*, Xuchuan Chen*, Dairu Liu*, Chenghuai Lin*, Yunrui Lian, Sikai Liang, Zhikai Zhang, Yu Guan, Jilong Wang, Wenyao Zhang, Xinqiang Yu, He Wang†, Li Yi†

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.

Paper Project Code

EMNLP

Evaluating Time Awareness and Cross-modal Active Perception of Large Models via 4D Escape Room Task

Yurui Dong*, Ziyue Wang*, Dairu Liu*, Shuyun Lu*, Xuechen Liu*, Fuwen Luo, Peng Li†, Yang Liu†

Conference on Empirical Methods in Natural Language Processing (EMNLP), 2026.

Paper Project

NeurIPS WS

Thinking with Visual Abstract: Enhancing Multimodal Reasoning via Visual Abstraction

Dairu Liu*, Ziyue Wang*, Minyuan Ruan*, Fuwen Luo, Chi Chen, Peng Li†, Yang Liu†

NeurIPS 2025 Workshop.

Paper Project

arXiv

Where is Physical RSI capability from? An Observational Scaling Law of LLM-as-policy

Zekai Li*, Jiayi Yu*, Jinlu Zhang*, Luyang Niu*, Zekun Qi*, Jun Zheng, Xuchuan Chen, Dairu Liu, Yintianrun Zhang, He Wang†, Li Yi†

arXiv, 2026.

Paper Project

ICML

LIMMT: Less is More for Motion Tracking

Yu Guan*, Zekun Qi*, Chenghuai Lin, Xuchuan Chen, Dairu Liu, Wenyao Zhang, Jilong Wang, Xinqiang Yu, He Wang†, Li Yi†

International Conference on Machine Learning (ICML), 2026.

Paper

Awards

• [2026] National Scholarship.
• [2024] National Scholarship.