Huanyu Zhang

Huanyu Zhang 张桓瑜

Ph.D Student

University of Chinese Academy of Sciences

Institute of Automation, Chinese Academy of Sciences

Intern @ Monnshot AI

About Me

I am a forth-year Ph.D Student of artificial intelligence at the State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS) in the Institute of Automation, Chinese Academy of Sciences (CASIA), advised by Prof.Tieniu Tan. Before starting my Ph.D, I was an undergraduate student in Automation at Xi’an Jiaotong University.

I am currently an intern at Moonshot AI, and I will be on the job market soon, actively seeking research and engineering opportunities in industry and academia.

I’m deeply passionate about advancing Multimodal AI, General AI, and Large Foundation Models, particularly through the integration of vision and text. My current research focuses on Visual Agents, and developing novel methodologies to unlock the intriguing and surprising potential of large models, while also having previously worked on multimodal reasoning. I warmly welcome academic collaborations and discussions—feel free to reach out!

Download CV

Interests

Multimodal Reasoning
Large Multimodal Models
Time Series Analysis

Education

PhD Student
CASIA
BSc in Automation
Xi'an Jiaotong University

Featured Publications

Multimodal Reasoning

Latent Sketchpad: Sketching Visual Thoughts to Elicit Multimodal Reasoning in MLLMs

Oct 29, 2025

Multimodal Reasoning

Scaling and Beyond: Advancing Spatial Reasoning in MLLMs Requires New Recipes

By dynamically retrieving relevant information from an external knowledge base, our TimeRAF enhances prediction accuracy, leading to more precise zero-shot forecasting performance.

Apr 21, 2025

Time Series Forecasting

TimeRAF: Retrieval-Augmented Foundation model for Zero-shot Time Series Forecasting (TKDE 2025)

By dynamically retrieving relevant information from an external knowledge base, our TimeRAF enhances prediction accuracy, leading to more precise zero-shot forecasting performance.

Dec 30, 2024

Time Series Classification

LogoRA: Local-Global Representation Alignment for Robust Time Series Classification (TKDE 2024)

We propose the Local-Global Representation Alignment framework (LogoRA), which combines multi-scale convolutional and transformer encoders, integrates representations with a fusion module, and employs advanced alignment strategies, achieving state-of-the-art performance on four time-series datasets.

Sep 30, 2024

Recent Publications

Huanyu Zhang*, Wenshan Wu*, Chengzu Li, Ning Shang, Yan Xia, Yangyu Huang, Yifan Zhang, Li Dong, Zhang Zhang, Liang Wang, Tieniu Tan, Furu Wei (2025). Latent Sketchpad: Sketching Visual Thoughts to Elicit Multimodal Reasoning in MLLMs. arXiv preprint arXiv:2510.24514.

PDF Cite

Huanyu Zhang*, Chengzu Li*, Wenshan Wu, Shaoguang Mao, Yifan Zhang, Haochen Tian, Ivan Vulić, Zhang Zhang, Liang Wang, Tieniu Tan, Furu Wei (2025). Scaling and Beyond: Advancing Spatial Reasoning in MLLMs Requires New Recipes. *arXiv preprint arXiv:2504.15037 *.

PDF Cite

Chengzu Li, Wenshan Wu, Huanyu Zhang, Yan Xia, Shaoguang Mao, Li Dong, Ivan Vulić, Furu Wei (2025). Imagine while Reasoning in Space: Multimodal Visualization-of-Thought. arXiv preprint arXiv:2501.07542.

PDF Cite

Yi-Fan Zhang, Huanyu Zhang, Haochen Tian, Chaoyou Fu, Shuangqing Zhang, Junfei Wu, Feng Li, Kun Wang, Qingsong Wen, Zhang Zhang, Others (2025). MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?. The Thirteenth International Conference on Learning Representations.

PDF Cite Code Dataset Project

Huanyu Zhang, Chang Xu, Yi-Fan Zhang, Zhang Zhang, Liang Wang, Jiang Bian, Tieniu Tan (2024). TimeRAF: Retrieval-Augmented Foundation model for Zero-shot Time Series Forecasting (TKDE 2025). IEEE Transactions on Knowledge and Data Engineering.

PDF Cite

See all publications