Biography

Hi there! I am currently a 3rd-year Ph.D. student at the Software Institute and the State Key Laboratory for Novel Software Technology, Nanjing University (NJU), fortunate to be co-advised by Prof. Tieke He and Prof. Jianhua Zhao. Prior to that, I received my B.Eng. degree (2024) in Software Engineering from Nanjing University, advised by Prof. Tieke He. You can find my CV online.

I am currently a Research Intern at Ant Group (i.e., Alipay, the largest fintech company in China), mentored by Dr. Tengfei Liu, focusing on tabular and relational foundation models. Previously, I also interned at Ant Group mentored by Dr. Yongchao Liu, working on learning with graphs and tabular data. I have published 10+ papers in top-tier AI and data mining conferences and journals, including NeurIPS, ICLR, KDD, WWW, ACL and TKDE.

Research Interests

  • Structured Data & LLMs: Exploring the synergy between large language models and structured data (graphs, knowledge graphs, and tabular data), pursuing two complementary directions: enhancing LLMs with structured data and LLM-assisted structured data learning
  • Tabular & Graph Foundation Models: Building foundation models from synthetic data via Prior-data Fitted Networks (PFNs) for in-context prediction across tables and graphs, iteratively refining the pipeline from prior design through synthetic data generation to foundation model training

Earlier in my research journey, I extensively explored scalable and trustworthy graph learning (including attributed graph clustering, graph anomaly detection, knowledge distillation, and self-supervised learning). An interactive synthesis of my first two Ph.D. years, tracing the trajectory from Attributed Graph Clustering to Graph Foundation Models, is available in English and 中文版. My long-term goal is to develop technically sound, scalable, and unified foundation models for structured data that deliver practical and impactful value in real-world applications.

News

  • 16 Oct 2026 Honored with the NeurIPS 2026 Top Reviewer, NeurIPS 2026 Financial Assistance Award, and China National Scholarship!
  • 25 Sep 2026 One paper on Attributed Graph Clustering accepted by NeurIPS 2026 (ED Track)!
  • 25 Sep 2026 One paper on LLM for Text-Attributed Graph Learning accepted by NeurIPS 2026!
  • 31 Jul 2026 Selected for the Postgraduate Research & Practice Innovation Program of Jiangsu Province with project Research on Semantic Hierarchy Discovery for Text-Attributed Graphs!
  • 22 Jun 2026 Won the Insta360 Think Bold Scholarship (Grand Prize), with sincere thanks to Insta360!
  • 06 Apr 2026 Three papers on Knowledge Graph Completion, RL for LLMs Survey, and Dynamic Graph Inference accepted by ACL 2026!
  • 10 Feb 2026 Released PyAGC, a comprehensive library and benchmark for Attributed Graph Clustering!
  • 26 Jan 2026 One paper on Graph Anomaly Detection accepted by ICLR 2026!
  • 13 Jan 2026 Two papers on (Foundational) Graph Anomaly Detection accepted by WWW 2026!
  • 24 Nov 2025 One paper on LLM for Text-Attributed Graph Learning accepted by KDD 2026!
  • 08 Nov 2025 Two papers on Entity Alignment and Knowledge Editing accepted by AAAI 2026!
  • 15 Oct 2025 Won the China National Scholarship for graduate students!
  • 17 Jul 2025 Excited to start my research internship at Ant Group, mentored by Dr. Yongchao Liu!
  • 13 Jul 2025 One paper on GNN-to-MLP Knowledge Distillation accepted by TKDE 2025!
  • 25 Jan 2025 One paper on GNN-to-MLP Knowledge Distillation accepted by DASFAA 2025!
  • 21 Jan 2025 One paper on Heterogeneous Text-Attributed Graph accepted by WWW 2025!
  • 21 Dec 2024 One paper on GAE for Link Prediction accepted by ICASSP 2025!
  • 24 Oct 2024 One paper on Graph Self-Supervised Learning accepted by Neural Networks 2024!
  • 20 Jul 2024 One paper on Attributed Graph Clustering accepted by TKDE 2024!
  • 16 Jul 2024 One paper on Attributed Graph Clustering accepted by CIKM 2024!
  • 27 May 2024 One paper on Graph Self-Supervised Learning accepted by ECML-PKDD 2024!

Publications

Projects

PyAGC: Attributed Graph Clustering Library

A modular library for Attributed Graph Clustering (AGC) built on PyTorch Geometric. It unifies 20+ algorithms under the Encode-Cluster-Optimize framework and supports mini-batch training on 111M+ nodes. Accompanied by a comprehensive survey bridging academia and industry perspectives.

StackOverflow-RoBERTa-base-Sentiment

A fine-tuned RoBERTa-base model for sentiment analysis on software engineering texts. It has garnered 3,108,129+ downloads on Hugging Face🤗 since June 2023.

Awesome Link Prediction

A curated and continuously updated collection of papers and resources on link prediction across graphs, temporal graphs, heterogeneous graphs, etc. Community contributions and stars are appreciated.

Heterogeneous Text-Attributed Graph Datasets

A collection of large, diverse benchmark datasets for heterogeneous text-attributed graphs (HTAGs), covering multiple domains such as movies, social, academia, literature, and patents.

Experience

Ant Group (i.e., Alipay, the largest fintech company in China)
Research Intern
Oct. 2026 - Present
Hangzhou, China

Research focus: tabular and relational foundation models; pre-train Transformer-based tabular FMs on synthetic tabular priors and real-world tabular data for in-context tabular prediction, reshaping industrial and scientific-discovery workflows that long rely on per-task training and tuning of gradient-boosted trees (e.g., XGBoost, LightGBM) and neural tabular models (e.g., RealMLP, TabM).

Ant Group (i.e., Alipay, the largest fintech company in China)
Research Intern
July 2025 - Feb. 2026
Hangzhou, China

Conducted research on learning with structured data, including graphs and tabular data.

  • LLMs for graphs: Developed an LLM-assisted framework to construct hierarchical label taxonomies for text-rich graphs and incorporate hierarchical knowledge into node representations. [KDD 2026]
  • Default prediction: Investigated pre-training on multi-source data to enhance user default prediction in Buy-Now-Pay-Later scenarios.
  • Graph anomaly detection: Adapted tabular foundation models by integrating graph structure into tabular features, enabling cross-graph anomaly detection via in-context learning without retraining. [WWW 2026]
  • Industrial benchmarking: Established a scalable, industrial-grade benchmark for attributed graph clustering and released the PyAGC codebase and toolkit for reproducible evaluation on large-scale graphs. [NeurIPS 2026] [GitHub]

Education

Nanjing University (NJU)
Ph.D. Student in Software Engineering
Sept. 2024 - Present
Nanjing, China
Nanjing University (NJU)
Bachelor in Software Engineering
Sept. 2020 - June 2024
Nanjing, China

Services

Program Committee Member / Reviewer

  • NeurIPS 2026 (Top Reviewer), KDD 2025&2026&2027, WWW 2027, PAKDD 2027, ICANN 2025&2026
  • Scientific Reports
  • Pattern Recognition (PR)
  • IEEE Transactions on Knowledge and Data Engineering (TKDE)

Teaching Assistant

  • Big Data Analytics (Fall 2024)
  • Data Management Fundamentals (Spring 2025)

Grants

  • Young Talent Support Program for Doctoral Students, CAST (China Association for Science and Technology), 2026 (中国科协青年人才托举工程博士生专项计划)
  • Postgraduate Research & Practice Innovation Program of Jiangsu Province, 2026 (江苏省研究生科研实践创新计划)

Honors

  • China National Scholarship, 2026
  • NeurIPS 2026 Financial Assistance Award
  • NeurIPS 2026 Top Reviewer
  • Insta360 Think Bold Scholarship (Grand Prize), 2026
  • China National Scholarship, 2025