Hi there Welcome to my Homepage!

Hi! I am a senior (4th-year) undergraduate student at Xidian University.

My research interests lie at the intersection of natural language processing and computer vision, with a particular focus on multimodal representation learning. My long-term goal is to understand how human-like intelligence can emerge and to develop unified models that connect the physical world with human cognition.

I enjoy exploring new research directions and collaborating with researchers across disciplines. In my spare time, I also enjoy exploring Buddhist philosophy. Feel free to reach out if you are interested in research collaborations or other opportunities.

One day, I’ll become a chef and open my own restaurant; if you have my contact info, you’ll get a discount. 😏

News

Education

Xidian University 2023.09 - Present

Undergraduate Student

B.E. at School of Telecommunication Engineering; advised by Xiumei Wang

Experience

Cardinal AI Lab, InfRec 2026.07 - Present

Chief Research Officer (Founding Member)

Leading research on recursive self-improvement and long-horizon world interaction.

Alibaba Group 2026.05 - 2026.07

MLLM Research Intern

Research on post-training, visual-semantic alignment, and large-scale e-commerce retrieval.

Institute of Artificial Intelligence of China Telecom 2026.01 - 2026.03

LLM Research Intern

AI infrastructure for LLM training and inference on Huawei Ascend 910B.

National University of Singapore 2024.10 - 2025.10

Research Assistant · HPC-AI Lab

Advised by Yang You, Wangbo Zhao and Pengfei Zhou

Publications

Core Publications Full Publications List

(* equal contribution · † corresponding author · ‡ project leader)

The Trinity: Capability-Driven Recursive Self-Improvement for Multimodal Agents
Weidong Tang^*, R. Wang^*, J. Zhang^*, H. Wang, H. Chen, Z. Zhang, Y. Ma, Z. Jiang, Z. Guo, H. Li, Z. Jiao, Z. Cui, X. Yin, S. Liang, Z. Wan, H. Lin, W. Zhao, K. Zhang, P. Zhou^†, O. Zhang^†, & Y. You^†.
We introduce a capability-driven framework that couples world generation with agent harness evolution, enabling multimodal agents to recursively improve and transfer learned capabilities to unseen tasks.
arXiv preprint, 2026   [Paper] [GitHub] [Project Page]

GroupToM-Bench: Benchmarking Group Theory of Mind and Nonlinear Social Emergence in MLLMs
Weidong Tang, Jierui Li, Yueling Hou, Zihan Mei, Can Zhang, Xinyan Wan, Zhiyuan Liang, Pengfei Zhou†, Yang You, Wangbo Zhao†
We propose GroupToM-Bench and show that current models fail at nonlinear group reasoning despite strong individual-level ToM, exposing a clear group cognitive gap.
ACL 2026   [Poster] [Paper] [Dataset]

MagicBench: Diagnosing Visual Agency Loss and Semantic Dependency in Multimodal LLMs
Tang Da Huang, Weidong Tang‡, Wen Qi Xu, Xianpeng Guo†
The authors introduce MagicBench, a video dataset of magic tricks, to test if Multimodal LLMs actually rely on visual physics or if they are overly dependent on language narratives (which, in magic tricks, are deliberately deceptive).
ACL 2026   [Paper] [code]

Efficient Video Object Segmentation and Tracking with Recurrent Dynamic Submodel
Weidong Tang, Zhiyuan Liang, Xinyan Wan, Chen Zhu, Zhaopan Xu, Pengfei Zhou, Yan Song, Yang You, Wangbo Zhao†
Proposed a Recurrent Dynamic Submodel for efficient Video Object Segmentation and Tracking. By integrating temporal-prior-guided global dynamic routing and Importance-aware LoRA, it achieves an optimal trade-off between performance and speed using minimal trainable parameters and training data.
CVPR 2026   [Poster] [Paper]

Inference-Time Scaling for Visual AutoRegressive Modeling by Searching Representative Samples
Weidong Tang, Xinyan Wan, Siyu Li, Xiumei Wang†
Explored inference-time scaling in discrete spaces by mapping them to continuous spaces to obtain density distributions, thereby optimizing the sampling of early coarse-scale features.
PRCV 2025   [Poster] [arXiv] [code]

Projects

WowPage
Weidong Tang, Yue Su.
In collaboration with Yue Su, I refined and improved his original homepage template. A clean standalone template version is coming soon.
Project   [code] [RedNote]

Awards

• 2025 & 2026
  Second-Class Academic Scholarship Xidian University · Ranked 5/99 (2025)

Talks

Services

• 2026.10 – Present
  Coordinator Xidian Graduate Admissions Experience-Sharing Group (3,000 members)
• 2025.06 – Present
  AI Team Lead 🌊Xidian–Inspur Club
• Present
  Reviewer ICIC 2026, PRCV (2025, 2026), ICLR 2027