Zeyang Zhang

Ph.D. Candidate, Computer Science and Technology
Tsinghua University
 zy-zhang20[at]mails[dot]tsinghua[dot]edu[dot]cn
 wondergo2017[at]gmail[dot]com

About me

Hi, I'm Zeyang! I'm currently a final-year Ph.D candidate in the Department of Computer Science & Technology at Tsinghua Univeristy, advised by Prof. Wenwu Zhu and Prof. Xin Wang. I got my Bachelor degree in Department of Computer Science and Technology at Tsinghua University in 2020. I also got my Second Degree in School of Economics and Management at Tsinghua University in 2020. I have published 20+ papers in prestigious conferences, e.g., NeurIPS, ICML, ACM KDD, ICLR, AAAI, CVPR, etc.

Research interests

My research interests lie in Open-environment Graph Machine Learning, including the subtopics:

Automated Architecture Adaptation: adapt to new environments, datasets, tasks by automatically designing, searching, modifying the neural network architectures
Out-of-distribution Generalization: generalize the neural network models under distribution shifts by discovering and exploiting the invariant patterns across environments
Large Language Models: enhance the zero/few-shot learning and reasoning abilities of the models in open environments by leveraging LLMs' strong abilities (textual understanding/planning/etc.)
Ai4Science Applications: solve the scientific problems by leveraging the AI techniques, including mathematical problems, drug discoveries, medical care, etc.

Selected papers

The full list of publications can be found in the publications page.
* indicates equal contributions.

Subtopic 1: Automated Architecture Adaptation

ICML’24

Disentangled Continual Graph Neural Architecture Search with Invariant Modular Supernet

Zeyang Zhang, Xin Wang, Yijian Qin, Hong Chen, Xu Chu, Ziwei Zhang, and Wenwu Zhu

In ICML, 2024

Bib

@inproceedings{gasim,
  title = {Disentangled Continual Graph Neural Architecture Search with Invariant Modular Supernet},
  author = {Zhang, Zeyang and Wang, Xin and Qin, Yijian and Chen, Hong and Chu, Xu and Zhang, Ziwei and Zhu, Wenwu},
  booktitle = {ICML},
  topic = {auto},
  year = {2024},
}

NeurIPS’23

Unsupervised Graph Neural Architecture Search with Disentangled Self-supervision

Zeyang Zhang, Xin Wang, Ziwei Zhang, Guangyao Shen, Shiqi Shen, and Wenwu Zhu

In Advances in Neural Information Processing Systems, 2023

Bib HTML

@inproceedings{zhang2023unsupervised,
  title = {Unsupervised Graph Neural Architecture Search with Disentangled Self-supervision},
  author = {Zhang, Zeyang and Wang, Xin and Zhang, Ziwei and Shen, Guangyao and Shen, Shiqi and Zhu, Wenwu},
  booktitle = {Advances in Neural Information Processing Systems},
  year = {2023},
  topic = {auto},
}

AAAI’23

Dynamic Heterogeneous Graph Attention Neural Architecture Search

Zeyang Zhang, Ziwei Zhang, Xin Wang, Yijian Qin, Zhou Qin, and Wenwu Zhu

In Thirty-Seventh AAAI Conference on Artificial Intelligence, 2023

Bib HTML

@inproceedings{zhang2023dynamic,
  title = {Dynamic Heterogeneous Graph Attention Neural Architecture Search},
  author = {Zhang, Zeyang and Zhang, Ziwei and Wang, Xin and Qin, Yijian and Qin, Zhou and Zhu, Wenwu},
  booktitle = {Thirty-Seventh {AAAI} Conference on Artificial Intelligence},
  year = {2023},
  topic = {auto},
}

Subtopic 2: Out-of-distribution Generalization

Arxiv

Out-of-Distribution Generalized Dynamic Graph Neural Network with Disentangled Intervention and Invariance Promotion

Zeyang Zhang, Xin Wang, Ziwei Zhang, Haoyang Li, and Wenwu Zhu

arXiv preprint arXiv:2311.14255, 2023

arXiv Bib

@article{zhang2023outofdistribution,
  title = {Out-of-Distribution Generalized Dynamic Graph Neural Network with Disentangled Intervention and Invariance Promotion},
  author = {Zhang, Zeyang and Wang, Xin and Zhang, Ziwei and Li, Haoyang and Zhu, Wenwu},
  year = {2023},
  journal = {arXiv preprint arXiv:2311.14255},
  topic = {ood},
}

NeurIPS’23

Spectral Invariant Learning for Dynamic Graphs under Distribution Shifts

Zeyang Zhang, Xin Wang, Ziwei Zhang, Zhou Qin, Weigao Wen, Hui Xue, Haoyang Li, and Wenwu Zhu

In Advances in Neural Information Processing Systems, 2023

Bib HTML

@inproceedings{zhang2023spectral,
  title = {Spectral Invariant Learning for Dynamic Graphs under Distribution Shifts},
  author = {Zhang, Zeyang and Wang, Xin and Zhang, Ziwei and Qin, Zhou and Wen, Weigao and Xue, Hui and Li, Haoyang and Zhu, Wenwu},
  booktitle = {Advances in Neural Information Processing Systems},
  year = {2023},
  topic = {ood},
}

NeurIPS’22

Dynamic graph neural networks under spatio-temporal distribution shift

Zeyang Zhang, Xin Wang, Ziwei Zhang, Haoyang Li, Zhou Qin, and Wenwu Zhu

In Advances in Neural Information Processing Systems, 2022

Bib HTML

@inproceedings{zhang2022dynamic,
  title = {Dynamic graph neural networks under spatio-temporal distribution shift},
  author = {Zhang, Zeyang and Wang, Xin and Zhang, Ziwei and Li, Haoyang and Qin, Zhou and Zhu, Wenwu},
  booktitle = {Advances in Neural Information Processing Systems},
  year = {2022},
  topic = {ood},
}

Subtopic 3: Large Language Models

KDD’24

LLM4DyG: Can Large Language Models Solve Spatial-Temporal Problems on Dynamic Graphs?

Zeyang Zhang, Xin Wang, Ziwei Zhang, Haoyang Li, Yijian Qin, and Wenwu Zhu

In ACM SIGKDD, 2024

Bib

@inproceedings{zhang2023LLM4DyG,
  title = {LLM4DyG: Can Large Language Models Solve Spatial-Temporal Problems on Dynamic Graphs?},
  author = {Zhang, Zeyang and Wang, Xin and Zhang, Ziwei and Li, Haoyang and Qin, Yijian and Zhu, Wenwu},
  booktitle = {ACM SIGKDD},
  year = {2024},
  topic = {llm},
}

NeurIPS’23w

Graph Meets LLMs: Towards Large Graph Models

Ziwei Zhang, Haoyang Li, Zeyang Zhang, Yijian Qin, Xin Wang, and Wenwu Zhu

In Conference on Neural Information Processing Systems Workshop, 2023

arXiv Bib

@inproceedings{zhang2023large,
  title = {Graph Meets LLMs: Towards Large Graph Models},
  author = {Zhang, Ziwei and Li, Haoyang and Zhang, Zeyang and Qin, Yijian and Wang, Xin and Zhu, Wenwu},
  booktitle = {Conference on Neural Information Processing Systems Workshop},
  year = {2023},
  topic = {llm},
}

Arxiv

Exploring the Potential of Large Language Models in Graph Generation

Yang Yao, Xin Wang, Zeyang Zhang, Yijian Qin, Ziwei Zhang, Xu Chu, Yuekui Yang, Wenwu Zhu, and Hong Mei

arXiv preprint arXiv:2403.14358, 2024

Bib

@article{yao2024exploring,
  title = {Exploring the Potential of Large Language Models in Graph Generation},
  author = {Yao, Yang and Wang, Xin and Zhang, Zeyang and Qin, Yijian and Zhang, Ziwei and Chu, Xu and Yang, Yuekui and Zhu, Wenwu and Mei, Hong},
  journal = {arXiv preprint arXiv:2403.14358},
  year = {2024},
  topic = {llm},
}

Subtopic 4: AI4Science Applications

BioRxiv
Decoding cell identity with multi-scale explainable deep learning

Jun Zhu*, Zeyang Zhang*, Yujia Xiang, Beini Xie, Xinwen Dong, Linhai Xie, Peijie Zhou, Rongyan Yao, Xiaowen Wang, Yang Li, Fuchu He, Wenwu Zhu, Ziwei Zhang, and Cheng Chang

bioRxiv, 2024

Abs Bib HTML

Cells are the fundamental structural and functional units of life. Studying the definition and composition of different cell types can help us understand the complex mechanisms underlying biological diversity and functionality. The increasing volume of extensive single-cell omics data makes it possible to provide detailed characterisations of cell types. Recently, there has been a rise in deep learning-based approaches that generate cell type labels solely through mapping query data to reference data. However, these approaches lack multi-scale descriptions and interpretations of identified cell types. Here, we propose Cell Decoder, a biological prior knowledge informed model to achieve multi-scale representation of cells. We implemented automated machine learning and post-hoc analysis techniques to decode cell identity. We have shown that Cell Decoder compares favourably to existing methods, offering multi-view interpretability for decoding cell identity and data integration. Furthermore, we have showcased its applicability in uncovering novel cell types and states in both human bone and mouse embryonic contexts, thereby revealing the multi-scale heterogeneity inherent in cell identities.Competing Interest StatementThe authors have declared no competing interest.
@article{Zhu2024.02.05.578922, author = {Zhu*, Jun and Zhang*, Zeyang and Xiang, Yujia and Xie, Beini and Dong, Xinwen and Xie, Linhai and Zhou, Peijie and Yao, Rongyan and Wang, Xiaowen and Li, Yang and He, Fuchu and Zhu, Wenwu and Zhang, Ziwei and Chang, Cheng}, title = {Decoding cell identity with multi-scale explainable deep learning}, elocation-id = {2024.02.05.578922}, year = {2024}, doi = {10.1101/2024.02.05.578922}, publisher = {Cold Spring Harbor Laboratory}, eprint = {https://www.biorxiv.org/content/early/2024/02/08/2024.02.05.578922.full.pdf}, journal = {bioRxiv}, topic = {ai4science}, }

MedAI’23

Out-of-Distribution Generalized Dynamic Graph Neural Network for Human Albumin Prediction

Zeyang Zhang*, Xingwang Li*, Fei Teng, Ning Lin, Xueling Zhu, Xin Wang, and Wenwu Zhu

In IEEE International Conference on Medical Artificial Intelligence, 2023

Bib HTML

@inproceedings{zhang2023ood,
  title = {Out-of-Distribution Generalized Dynamic Graph Neural Network for Human Albumin Prediction},
  author = {Zhang*, Zeyang and Li*, Xingwang and Teng, Fei and Lin, Ning and Zhu, Xueling and Wang, Xin and Zhu, Wenwu},
  booktitle = {IEEE International Conference on Medical Artificial Intelligence},
  year = {2023},
  topic = {ai4science},
}

AAAI’22

Learning to Solve Travelling Salesman Problem with Hardness-adaptive Curriculum

Zeyang Zhang, Ziwei Zhang, Xin Wang, and Wenwu Zhu

Thirty-Fifth AAAI Conference on Artificial Intelligence, 2022

Bib HTML

@article{zhang2022learning,
  title = {Learning to Solve Travelling Salesman Problem with Hardness-adaptive Curriculum},
  author = {Zhang, Zeyang and Zhang, Ziwei and Wang, Xin and Zhu, Wenwu},
  journal = {Thirty-Fifth {AAAI} Conference on Artificial Intelligence},
  year = {2022},
  topic = {ai4science},
}

Projects

AutoGL: A toolkit and platform towards automatic machine learning on graphs.
Paper collections for the topic of large graph models.

Honors

Tsinghua Comprehensive Excellence Scholarship (2019, 2022, 2023)
Outstanding Consouller in THU CS (2023)
Tsinghua Longhu Scholarship (2023)
Tsinghua Outstanding Student Leadership Award (2019, 2022)
Tsinghua '84' Future Innovation Scholarship (2022)
Outstanding Graduate (Bachelor) at THU CS (2020)
Shanghai Junyuan Scholarship (2015, 2017, 2019)
Excellent Academic Award in Tsinghua University (2018)
First Prize in National Physics Competition for Undergraduate in Some Region (2017)

Internship

(2021.11 ~ 2022.06) Alibaba Group, researching on graph algorithms for e-commerce risk management. Our paper is published in NeurIPS'22.
(2022.07 ~ 2022.09) Tencent Wechat, researching on online video tagging and recommendation systems. Our paper is published in NeurIPS'23.
(2022.10 ~ 2024.02) Alibaba Group, researching on OOD generalization and LLMs for online transactions. Our paper is published in AAAI'23 and NeurIPS'23.

Professional Services

Conference reviewer / Program committee: NeurIPS (2022, 2023, 2024), ICLR (2023, 2024), AAAI (2023, 2024), WWW (2023, 2024), CVPR (2024), ICML (2024), KDD(2024), ACM MM (2024)
Journal reviewer: IEEE Transactions on Neural Networks and Learning Systems (TNNLS), IEEE Transactions on Visualization and Computer Graphics (TVCG), Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
Teaching assistant: Introduction to Data Science, Tsinghua University (Fall 2020, 2021, 2022, 2023)
Consouller in Department of Compuster Science and Technology, Tsinghua University (2020, 2021, 2022, 2023)