Currently, Xunkai Li is a fourth year Ph.D. student of School of Computer Science in Beijing Institute of Technology supervised by Prof. Rong-Hua Li. Prior to that, he received his B.Eng. degree in computer science department from Shandong University in June 2022. As an active member of the research team, Xunkai is the organizer or the contributor of scalable graph learning system project TKDE 2025 SGL, federated graph learning system project VLDB 2025 OpenFGL, graph unlearning system project NeurIPS 2025 OpenGU, directed graph learning system project TKDE 2026 DcDGL, and multimodal graph learning system project ICML 2026 OpenMAG. Meanwhile, he is the Principal Investigator of the National Natural Science Foundation Youth Student Basic Research Project (2025) and is selected for the China Association for Science and Technology (CAST) Young Talent Support Program (2025).

Email: cs.xunkai.li@gmail.com

Wechat (微信): 18045124943

🌟 🌟 If you are interested in collaborating with me or want to have a chat, always feel free to contact me through e-mail or WeChat ：）

My research interests center on Data-centric Graph Intelligence, including Graph Database Theory and Machine Learning, Multimodal Understanding and Generation, and AI4Science, with a focus on constructing and mining graph-structured data that contains deep relational and causal knowledge from emerging multimodal domains. The goal of my work is to bridge the gap between theoretical advancements and practical applications in real-world industrial and scientific scenarios.

Graph Database Theory and Machine Learning: In an era where everything is interconnected, graphs have become a fundamental tool for modeling complex relationships among entities, playing an important role in domains such as social networks (relationships), recommendation systems (preferences), and molecular (atoms and chemical bonds). Theoretical studies on graph databases provide essential foundations for efficient relational data management. Building upon this foundation, graph machine learning, especially graph neural networks (GNNs), further enhances the ability to mine structured data through AI-driven techniques, offering more efficient solutions and promoting the development of related fields. In this context, I am particularly interested in Graph Management and ML in the era of LLM.

Multi-Agent Systems: This key technology has emerged as a vital paradigm for eliciting the collective intelligence of large models, demonstrating remarkable efficacy in areas such as complex reasoning, automated code generation, and embodied decision-making. Existing research in this domain strives to establish a foundational framework for efficient agent collaboration, experience reuse, and autonomous system evolution. Building upon these advancements, graphs—with their inherent capacity to model and represent intricate relationships—offer highly effective solutions for optimizing multi-agent communication, memory management, and iterative skill acquisition. Currently, my research primarily focuses on Graph-driven Multi-Agent Systems.

Multimodal Understanding and Generation: In the multimodal attributed graphs we construct, nodes represent multimodal semantic entities, while edges capture their explicit or implicit contextual relationships. This new form of graph-structured data breaks the unnecessary isolation among entities in traditional multimodal learning and plays an important role in cross-modal retrieval and content generation. Based on this, multimodal GNNs leverage structured contexts to overcome communication barriers among entities, providing a new paradigm for cross-entity modal interaction and understanding. In this context, my research focuses on Structure-driven Multimodal Understanding and Generation.

AI4Science: The transformative potential of AI4Science lies in its capacity to revolutionize our understanding and transformation of the world, with profound implications across all aspects of human life. My goal is to advance the practical application of artificial intelligence in scientific domains, driving a new wave of innovation—particularly in Graph and Multimodal Intelligence for Drug, Protein, and Cell.

🔥 News

• 2026-4: seven paper are accepted by ICML 2026.
• 2026-4: One paper is accepted by TKDE 2026.
• 2026-4: One paper is accepted by TNNLS 2026.
• 2026-4: One paper is accepted by ICMR 2026.
• 2026-4: One paper is accepted by ACL 2026.
• 2026-4: One paper is accepted by ICME 2026.
• 2026-1: One paper is accepted by WWW 2026.
• 2026-1: One paper is accepted by TKDE 2026.
• 2025-12: One paper is accepted by TNNLS 2026.
• 2025-11: One paper is accepted by AAAI 2026.
• 2025-09: Three paper are accepted by NeurIPS 2025.
• 2025-08: Two paper are accepted by CIKM 2025.
• 2025-05: One paper is accepted by ICML 2025.
• 2025-04: One paper is accepted by IJCAI 2025.
• 2025-01: One paper is accepted by TKDE 2025.
• 2025-01: One paper is accepted by WWW 2025.
• 2025-01: One paper is accepted by VLDB 2025.
• 2024-08: One paper is accepted by VLDB 2025.
• 2024-04: One paper is accepted by IJCAI 2024.
• 2024-03: One paper is accepted by ICDE 2024.
• 2024-02: One paper is accepted by VLDB 2024.
• 2024-01: One paper is accepted by WWW 2024.
• 2023-12: One paper is accepted by AAAI 2024.
• 2023-10: One paper is accepted by ICDE 2024.
• 2023-08: One paper is accepted by VLDB 2023.

📝 Publications

First/Co-first Author

# indicates equal contribution, @ indicates corresponding author

ICML 2026

🎯 When LLMs Encounter Open-world Graph Learning: A Fresh View on Unlabeled Data Uncertainty, [Code]

Yanzhe Wen#, Xunkai Li#, Qi Zhang, Lei Zhu, Guang Zeng, Zhihan Zhang, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

ICML 2026

🎯 Toward Effective Multimodal Graph Foundation Model: A Divide-and-Conquer Based Approach, [Code]

Sicheng Liu#, Xunkai Li#, Daohan Su, Ru Zhang, Hongchao Qin, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

ICML 2026

🎯 HIAL: Towards Semantics-Aware Hypergraph Active Learning via Dual-Perspective Information Maximization, [Code]

Yanheng Hou#, Xunkai Li#, Yanzhe Wen, Zhenjun Li, Bing Zhou, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

ICML 2026

🎯 OpenMAG: A Comprehensive Benchmark for Multimodal-Attributed Graph, [Code]

Chenxi Wan#, Xunkai Li#, Yilong Zuo, Haokun Deng, Sihan Li, Bowen Fan, Hongchao Qin, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

TKDE 2026

🎯 Towards Data-centric Machine Learning on Directed Graphs: a Survey, [Code]

Henan Sun#, Xunkai Li#, Daohan Su, Junyi Han, Rong-Hua Li, Guoren Wang

• IEEE Transactions on Knowledge and Data Engineering (TKDE) 2026, CCF-A.

AAAI 2026

🎯 PAGE: A Unified Approach for Federated Graph Unlearning, [Code]

Yuming Ai#, Xunkai Li#, Jiaqi Chao, Bowen Fan, Zhengyu Wu, Yinlin Zhu, Rong-Hua Li, Guoren Wang

• Association for the Advancement of Artificial Intelligence (AAAI), 2026, CCF-A.

NeurIPS 2025

🎯 Towards Effective Federated Graph Foundation Model via Mitigating Knowledge Entanglement, [Code]

Yinlin Zhu#, Xunkai Li#, Jishuo Jia, Miao Hu, Di Wu, Meikang Qiu

• Neural Information Processing Systems (NeurIPS), 2025, CCF-A.

ICML 2025

🎯 Toward Data-centric Directed Graph Learning: An Entropy-driven Approach, [Code]

Xunkai Li, Zhengyu Wu, Kaichi Yu, Hongchao Qin, Guang Zeng, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2025, CCF-A.

WWW 2025

🎯 Toward Effective Digraph Representation Learning: A Magnetic Adaptive Propagation based Approach, [Code]

Xunkai Li, Daohan Su, Zhengyu Wu, Guang Zeng, Hongchao Qin, Rong-Hua Li, Guoren Wang

• The Web Conference (WWW), 2025, CCF-A.

VLDB 2025

🎯 OpenFGL: A Comprehensive Benchmark for Federated Graph Learning, [Code]

Xunkai Li, Yinlin Zhu, Boyang Pang, Guochen Yan, Yeyu Yan, Zening Li, Zhengyu Wu, Wentao Zhang, Rong-Hua Li, Guoren Wang

• International Conference on Very Large Data Bases (VLDB), 2025, CCF-A.

ICDE 2024

🎯 AdaFGL: A New Paradigm for Federated Node Classification with Topology Heterogeneity, [Code]

Xunkai Li, Zhengyu Wu, Wentao Zhang, Henan Sun, Rong-Hua Li, Guoren Wang

• IEEE International Conference on Data Engineering (ICDE), 2024, CCF-A.

AAAI 2024

🎯 Towards Effective and General Graph Unlearning via Mutual Evolution, [Code]

Xunkai Li#, Yulin Zhao#, Zhengyu Wu, Wentao Zhang, Rong-Hua Li, Guoren Wang

• Association for the Advancement of Artificial Intelligence (AAAI), 2024, CCF-A.
• 🎉🎉 Oral Presentation

WWW 2024

🎯 Rethinking Node-wise Propagation for Large-scale Graph Learning, [Code]

Xunkai Li, Jingyuan Ma, Zhengyu Wu, Daohan Su, Wentao Zhang, Rong-Hua Li, Guoren Wang

• The Web Conference (WWW), 2024, CCF-A.
• 🎉🎉 Oral Presentation

VLDB 2024

🎯 LightDiC: A Simple yet Effective Approach for Large-scale Digraph Representation Learning, [Code]

Xunkai Li, Meihao Liao, Zhengyu Wu, Daohan Su, Wentao Zhang, Rong-Hua Li, Guoren Wang

• International Conference on Very Large Data Bases (VLDB), 2024, CCF-A.

ICDE 2024

🎯 Breaking the Entanglement of Homophily and Heterophily in Semi-supervised Node Classification, [Code]

Henan Sun#, Xunkai Li#, Zhengyu Wu, Daohan Su, Rong-Hua Li, Guoren Wang

• IEEE International Conference on Data Engineering (ICDE), 2024, CCF-A.

VLDB 2023

🎯 FedGTA: Topology-aware Averaging for Federated Graph Learning, [Code]

Xunkai Li, Zhengyu Wu, Wentao Zhang, Yinlin Zhu, Rong-Hua Li, Guoren Wang

• International Conference on Very Large Data Bases (VLDB), 2023, CCF-A.

NCAA 2023

Effective Hybrid Graph and Hypergraph Convolution Network for Collaborative Filtering [code]

Xunkai Li#, Ronghui Guo#, Jianwen Chen, Youpeng Hu, Meixia Qu, Bin Jiang

• Neural Computing & Applications (NCAA) 2023, CCF-C.

Co-author

🎯 MagicDock: Toward Docking-oriented De Novo Ligand Design via Gradient Inversion, [Code]

Zekai Chen, Xunkai Li, Sirui Zhang, Henan Sun, Jia Li, Zhenjun Li, Bing Zhou, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

🎯 DANCE: Dynamic, Available, Neighbor-gated Condensation for Federated Text-Attributed Graphs, [Code]

Zekai Chen#, Haodong Lu#, Xunkai Li, Henan Sun, Jia Li, Hongchao Qin, Rong-Hua Li, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

🎯 Scalable Topology-Preserving Graph Coarsening with Graph Collapse, [Code]

Xiang Wu, Rong-Hua Li, Xunkai Li, Kangfei Zhao, Hongchao Qin, Guoren Wang

• International Conference on Machine Learning (ICML), 2026, CCF-A.

🎯 Towards Unbiased Federated Graph Learning: Label and Topology Perspectives, [Code]

Zhengyu Wu, Boyang Pang, Xunkai Li, Yinlin Zhu, Daohan Su, Bowen Fan, Rong-Hua Li, Guoren Wang, Chenghu Zhou

• IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2026, CCF-B.

🎯 Toward General and Robust LLM-enhanced Text-attributed Graph Learning, [Code]

Zihao Zhang, Xunkai Li@, Rong-Hua Li, Bing Zhou, Zhenjun Li, Guoren Wang

• International Conference on Multimedia Retrieval (ICMR), 2026, CCF-B.

🎯 AlgBench: To What Extent Do Large Reasoning Models Understand Algorithms?, [Code]

Henan Sun#, Kaichi Yu#, Yuyao Wang, Bowen Liu, Xunkai Li, Rong-Hua Li, Nuo Chen, Jia Li

• The Association for Computational Linguistics (ACL), 2026, CCF-A.

🎯 Toward Model-centric Heterogeneous Federated Graph Learning: A Knowledge-driven Approach, [Code]

Zhengyu Wu, Guang Zeng, Huilin Lai, Daohan Su, Jishuo Jia, Yinlin Zhu, Xunkai Li, Rong-Hua Li, Guoren Wang, Chenghu Zhou

• International Conference on Multimedia and Expo (ICME), 2026, CCF-B.

🎯 Unveiling the Vulnerability of Graph-LLMs: An Interpretable Multi-Dimensional Adversarial Attack on TAGs, [Code]

Bowen Fan#, Zhilin Guo#, Xunkai Li, Yihan Zhou, Bing Zhou, Zhenjun Li, Rong-Hua Li, Guoren Wang

• The Web Conference (WWW), 2026, CCF-A.

🎯 Effective Bug Detection in Graph Database Engines: An LLM-based Approach, [Code]

Jiayi Wu, Zhengyu Wu, Xunkai Li, Hongchao Qin, Ronghua Li, Guoren Wang

• IEEE Transactions on Knowledge and Data Engineering (TKDE) 2026, CCF-A.

🎯 ScaDyG: A New Paradigm for Large-scale Dynamic Graph Learning, [Code]

Xiang Wu, Xunkai Li, Rong-Hua Li, Kangfei Zhao, Guoren Wang

• IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2026, CCF-B.

🎯 GraphMaster: Automated Graph Synthesis via LLM Agents in Data-Limited Environments, [Code]

Enjun Du, Xunkai Li, Tian Jin, Zhihan Zhang, Rong-Hua Li, Guoren Wang

• Neural Information Processing Systems (NeurIPS), 2025, CCF-A.

🎯 OpenGU: A Comprehensive Benchmark for Graph Unlearning, [Code]

Bowen Fan, Yuming Ai, Xunkai Li, Zhilin Guo, Rong-Hua Li, Guoren Wang

• Neural Information Processing Systems (NeurIPS), 2025, CCF-A.

🎯 DiRW: Path-Aware Digraph Learning for Heterophily, [Code]

Daohan Su, Xunkai Li, Zhenjun Li, Yinping Liao, Rong-Hua Li, Guoren Wang

• Conference on Information and Knowledge Management (CIKM), 2025, CCF-B.

🎯 FedC4: Graph Condensation Meets Client-Client Collaboration for Efficient and Private Federated Graph Learning, [Code]

Zekai Chen, Xunkai Li, Yinlin Zhu, Rong-Hua Li, Guoren Wang

• Conference on Information and Knowledge Management (CIKM), 2025, CCF-B.

🎯 Rethinking Federated Graph Learning: A Data Condensation Perspective, [Code]

Hao Zhang, Xunkai Li, Yinlin Zhu, Lianglin Hu

• International Joint Conference on Artificial Intelligence (IJCAI), 2025, CCF-A.

🎯 Acceleration Algorithms in GNNs: A Survey

Lu Ma, Zeang Sheng, Xunkai Li, Xinyi Gao, Zhezheng Hao, Ling Yang, Wentao Zhang, Bin Cui

• IEEE Transactions on Knowledge and Data Engineering (TKDE) 2025, CCF-A.

🎯 Topology-preserving Graph Coarsening: An Elementary Collapse-based Approach, [Code]

Yuchen Meng, Rong-Hua Li, Longlong Lin, Xunkai Li, Guoren Wang

• International Conference on Very Large Data Bases (VLDB), 2025, CCF-A.

🎯 FedTAD: Topology-aware Data-free Knowledge Distillation for Subgraph Federated Learning, [Code]

Yinlin Zhu, Xunkai Li, Zhengyu Wu, Di Wu, Miao Hu, Rong-Hua Li

• International Joint Conference on Artificial Intelligence (IJCAI), 2024, CCF-A.

🎯 Adaptive Hypergraph Auto-Encoder for Relational Data Clustering [code]

Youpeng Hu, Xunkai Li, Yujie Wang, Yixuan Wu, Yining Zhao, Chenggang Yan, Jian Yin, Yue Gao

• IEEE Transactions on Knowledge and Data Engineering (TKDE) 2023, CCF-A.

🎯 FedASU: Attention-guided sensitivity unlearning framework for multi-scenario federated graph unlearning

Jiaqi Chao, Xunkai Li, Yuming Ai, Meixia Qu

• Expert Systems with Applications (ESWA) 2026, CCF-B.

🎯 Towards Effective Few-Shot OOD Detection for Text-Attributed Graphs via Topology-Text Consensus Modeling

Xu Wang, Yinlin Zhu, Xunkai Li, Meixia Qu, Wenyu Wang

• Knowledge-Based Systems (KBS) 2026, CCF-B.

🎯 FedSA-GCL: A Semi-Asynchronous Federated Graph Learning Framework with Personalized Aggregation and Cluster-Aware Broadcasting

Zhongzheng Yuan, Lianshuai Guo, Xunkai Li, Yinlin Zhu, Wenyu Wang, Meixia Qu

• Knowledge-Based Systems (KBS) 2026, CCF-B.

🎯 FedPPD: Towards Effective Subgraph Federated Learning via Pseudo Prototype Distillation

Qi Lin, Jishuo Jia, Yinlin Zhu, Xunkai Li, Bin Jiang, Meixia Qu

• Neural Networks (NN) 2025, CCF-B.

🎯 LoyalDE: Improving The Performance of Graph Neural Networks with Loyal Node Discovery and Emphasis

Haotong Wei, Yinlin Zhu, Xunkai Li, Bin Jiang

• Neural Networks (NN) 2023, CCF-B.

🎯 Handling Information Loss of Graph Convolutional Networks in Collaborative Filtering

Xin Xiong, Xunkai Li, Youpeng Hu, Yixuan Wu, Jian Yin

• Information Systems (IS) 2022, CCF-B.

🎯 Graph Relation Embedding Network for Click-through Rate Prediction

Yixuan Wu, Youpeng Hu, Xin Xiong, Xunkai Li, Ronghui Guo, Shuiguang Deng

• Knowledge and Information Systems (KAIS) 2022, CCF-B.

📖 Educations

• 2022.09 - 2028.06 (expected), Ph.D. in Computer Science, Beijing Institute of Technology (BIT)

• 2018.09 - 2022.06, B. Eng in Computer Science, Shandong University (SDU)

🧑‍💻 Service

• Conference Reviewers:
  • ICML, ICLR, NeurIPS, KDD, WWW, AISTATS, AAAI, IJCAI
• Journal Reviewers:
  • TKDE, TKDD, TNNLS, TBD

💻 Internships

• Research Intern
  • Company/Institution: Ant Group, Alibaba, Beijing.
  • Advisor: Researcher Ningtao Wang
  • Employment period: From 04/2026 to the present
• Research Intern
  • Company/Institution: The Hong Kong University of Science and Technology, Guangzhou.
  • Advisor: Prof. Jia Li
  • Employment period: From 2/2025 to the present
• Research Intern
  • Company/Institution: Ant Group, Alibaba, Beijing.
  • Advisor: Researcher Guang Zeng
  • Employment period: From 12/2024 to 08/2025
• Research Intern
  • Company/Institution: Large Language Model Center, IAAR, Shanghai.
  • Advisor: Dr. Zhiyu Li
  • Employment period: From 06/2024 to 12/2024
• Research Intern
  • Company/Institution: DCML Group, Peking University
  • Advisor: Prof. Wentao Zhang
  • Employment period: From 10/2021 to the present
• Research Intern
  • Company/Institution: iMoon Lab, Tsinghua University
  • Advisor: Prof. Yue Gao
  • Employment period: From 06/2020 to 06/2021