I got the Ph.D. degree from Zhejiang University, advised by Prof. Siliang Tang (汤斯亮). I am currently a Junior Researcher at Shanghai AI Lab, working with Dr. Lijun Wu and Dr. Conghui He. My current research focuses on post-training for vision-language models (VLMs), multimodal data valuation, and agentic reinforcement learning.

More specifically, I study how to assess the value of multimodal data, identify and construct high-value datasets for VLM post-training, and improve model capabilities through agentic RL. I am particularly interested in building scalable data-centric pipelines that better align multimodal foundation models with real-world reasoning and decision-making tasks. More details can be found on our project pages, Open Data Arena and SciVerse.

Please feel free to contact me if you are interested in internships or collaborations in these areas.

🔥 News

🗃️ Publications

“*” denotes equal contribution, “†” denotes corresponding author

Graph-Related Studies

🕸 Graph Generalization

WWW-25

GraphCLIP: Enhancing Transferability in Graph Foundation Models for Text-Attributed Graphs
Yun Zhu, Haizhou Shi, Xiaotang Wang, Yongchao Liu, Yaoke Wang, Boci Peng, Chuntao Hong, Siliang Tang†

KDD-25

Graph Triple Attention Network: A Decoupled Perspective
Xiaotang Wang*, Yun Zhu*†, Haizhou Shi, Yongchao Liu, Chuntao Hong

WWW-24

MARIO: Model Agnostic Recipe for Improving OOD Generalization of Graph Contrastive Learning
Yun Zhu, Haizhou Shi, Zhenshuo Zhang, Siliang Tang†

🔧 Graph Adaption

EMNLP-24 (Main)

Bridging Local Details and Global Context in Text-Attributed Graphs
Yaoke Wang*, Yun Zhu*, Wenqiao Zhang, Yueting Zhuang, Yunfei Li, Siliang Tang†

IJCAI-24

Efficient Tuning and Inference for Large Language Models on Textual Graphs
Yun Zhu*, Yaoke Wang*, Haizhou Shi, Siliang Tang†

WWW-24 (Oral)

GraphControl: Adding Conditional Control to Universal Graph Pre-trained Models for Graph Domain Transfer Learning
Yun Zhu*, Yaoke Wang*, Haizhou Shi, Zhenshuo Zhang, Siliang Tang†

LLM-Related Studies

🚀 Efficient Inference

IJCAI-23

SmartBERT: A Promotion of Dynamic Early Exiting Mechanism for Accelerating BERT Inference
Boren Hu*,Yun Zhu*, Jiacheng Li, Siliang Tang†

🔧 Post-Training

WWW-25-Industry-Track

Transferable and Forecastable User Targeting Foundation Model
Bin Dou, Baokun Wang†, Yun Zhu, et al.

KDD-24-Workshop

LLaSA: Large Language and E-Commerce Shopping Assistant
Shuo Zhang, Boci Peng, Xinping Zhao, Boren Hu, Yun Zhu, Yanjia Zeng, Xuming Hu†

💾 GraphRAG & Reflection

Under Review

Graph Retrieval-Augmented Generation: A Survey
Boci Peng*, Yun Zhu*, Yongchao Liu, Xiaohe Bo, Haizhou Shi, Chuntao Hong, Yan Zhang†, Siliang Tang

ACL-25 (Main)

Meta-Reflection: A Feedback-Free Reflection Learning Framework
Yaoke Wang, Yun Zhu, Xintong Bao, Wenqiao Zhang†, Suyang Dai, Kehan Chen†, Wenqiang Li, Gang Huang, Siliang Tang, Yueting Zhuang

💼 Internships

• 2024.2 - 2025.5: Research intern at AntGroup focusing on Foundation Models, Post-training and GraphRAG, mentored by Yongchao Liu (刘永超)

🎖 Honors and Awards

• 2025 Provincial Outstanding Graduate (PhD)
• 2023-2024 Award of Honor for Graduate
• 2022-2023 Honor for Graduates-Excellence in Academic (Practice) Innovation
• 2021-2022 Award of Honor for Graduate
• 2020-2021 Award of Honor for Graduate
• 2019-2020 Outstanding Graduate of Zhejiang Province, Provincial Government Scholarship, The First Prize Scholarship
• 2018-2019 Provincial Government Scholarship, The First Prize Scholarship
• 2017-2018 Provincial Government Scholarship, The First Prize Scholarship
• 2016-2017 Provincial Government Scholarship, The First Prize Scholarship

🎓 Educations

• 2020.09 - 2025.06, Ph.D., Zhejiang University, Hangzhou.
• 2016.09 - 2020.06, Undergraduate, Software Engineering, Zhejiang Univeristy of Technology, Hangzhou.
• 2013.09 - 2016.06, Xin’anjiang High School, Jiande, Hangzhou, Zhejiang.

📝 Blogs

• A Glance of Graph Transformers (Chinese)
• The Applications of Mutal Information in Deep Learning (Chinese)
• Gumbel-Softmax for Differentiable Sampling (Chinese)
• Graph OOD Generalization (Chinese)
• More blogs can be found in here

📚 Collections

• Awesome GNN works on NeurIPS2022
• GNN+LLM works submitted to ICLR2024
• Highlight GNN works on ICLR2023
• Awesome GNN works on ICML2022