I’m a Fourth-year Ph.D. candidate in the Department of Computer Science and Engineering, Washington University in St.Louis (WashU). I’m fortunately supervised by Dr. Yixin Chen and also closely work with Dr. Fuhai Li. My research primarily focuses on graph representation learning. Particularly, most of my research targets an ultimate goal: to design and train a graph foundation model (GFMs) with graph neural networks (GNNs). Specifically:
- Understanding and improving the expressiveness and structure learning ability of GNNs on graphs.
- Integrate GNNs with LLMs to design the architecture of GFMs and enable a single model to solve graph tasks in multiple levels (node/link/graph) and domains (citation/molecule/e-commercial).
- Design unified self-supervised pre-training tasks and graph prompting mechanism on graph domain to enable zero-shot transfer ability in GFMs.
- Collect and create large-scale graph datasets for the training of GFMs.
Besides that, I am also interested in the downstream application of GNNs in various domains like recommendation, precision medicine, planning, and reasoning.
I am actively looking for research internship in summer 2025. If you are interested in my work, feel free to chat or shoot me email for discussing any potential opportunities.
🔥 News
- 2024.09: GNN4TaskPlan is accepted by NeurIPS 2024, Congratulations to Xixi and Yifei!
- 2024.08: Check out our newest work on joint modeling of graph and language (paper,code). In this work, we propose GOFA, which interleave GNN layers into LLMs to enable LLM with the ability to reason on graph. We also design multiple novel large-scale unsupervised pretraining tasks for GOFA. GOFA achieves SOTA results across multiple benchmarking datasets!
- 2024.06: We release TAGLAS, an atlas of text-attributed graph datasets. We provide easy-to-use APIs for loading datasets, tasks, and evaluation metrics. The technical report is available in arxiv. The project is still in development, any suggestions are welcome.
- 2024.04: 🎉🎉 PathFinder is accepted by Frontiers in Cellular Neuroscience!
- 2024.01: 🎉🎉 COLA is accepted by WWW 2024! Congratulations to Hao!
- 2024.01: 🎉🎉 OFA is accepted by ICLR 2024 as a Spotlight(5%)!
- 2024.01: 🎉🎉 sc2MeNetDrug is accepted by PLOS Computational Biology!
- 2023.10: We have developed a novel R Shiny application sc2MeNetDrug for the analysis of single-cell RNA-seq data. This application enables the identification of activated pathways, up-regulated ligands and receptors, cell-cell communication networks, and potential drugs to inhibit dysfunctional networks. Moreover, it provides user-friendly UI for easy usage! Check out our GitHub repository and website for more details. This project is still ongoing, and we welcome any comments or suggestions!
- 2023.10: Leveraging the power of language and LLMs, we propose One-for-ALL (OFA), which is the first general framework that can use a single graph model to address (almost) all different graph classification tasks from different domains. Check out our preprint and code!
- 2023.09: 🎉🎉 (k,t)-FWL+, MAG-GNN, and d-DRFWL2(spotlight) are accepted by NeurIPS 2023!
- 2023.06: 🎉🎉 Passed the oral exam!
- 2022.09: 🎉🎉 Our paper “How powerful are K-hop message passing graph neural networks” is accepted by NeurIPS 2022. See you in New Orleans!
- 2022.08: 🎉🎉 Our paper “Reward delay attacks on deep reinforcement learning” is accepted by GameSec 2022.
📝 Selected Publications
One for All: Towards Training One Graph Model for All Classification Tasks
Hao Liu*, Jiarui Feng*, Lecheng Kong*, Ningyue Liang, Dacheng Tao, Yixin Chen, Muhan Zhang (* Equal contribution)
Graph Contrastive Learning Meets Graph Meta Learning: A Unified Method for Few-shot Node Tasks
Hao Liu, Jiarui Feng, Lecheng Kong, Dacheng Tao, Yixin Chen, Muhan Zhang
Extending the Design Space of Graph Neural Networks by Rethinking Folklore Weisfeiler-Lehman
Jiarui Feng, Lecheng Kong, Hao Liu, Dacheng Tao, Fuhai Li, Muhan Zhang, Yixin Chen
Distance-Restricted Folklore Weisfeiler-Leman GNNs with Provable Cycle Counting Power
Junru Zhou, Jiarui Feng, Xiyuan Wang, Muhan Zhang
MAG-GNN: Reinforcement Learning Boosted Graph Neural Network
Lecheng Kong, Jiarui Feng, Hao Liu, Dacheng Tao, Yixin Chen, Muhan Zhang
How powerful are K-hop message passing graph neural networks
Jiarui Feng, Yixin Chen, Fuhai Li, Anindya Sarkar, Muhan Zhang
Reward Delay Attacks on Deep Reinforcement Learning
Anindya Sarkar, Jiarui Feng, Yevgeniy Vorobeychik, Christopher Gill, Ning Zhang
You can browse my full publication list in Google Scholar
🎖 Honors and Awards
- 2023.10 NeurIPS 2023 Travel Award.
- 2021.07 ICIBM 2021 Travel Award.
📖 Educations
- 2021.09 - Present, Ph.D student, Washington University in St. Louis, MO, USA.
- 2019.09 - 2021.05, Master, Washington University in St. Louis, MO, USA.
- 2015.09 - 2019.06, Bachelor, South China Univerity of Technology, GuangZhou, China.
💻 Internships
- 2024.06 - 2024.09 Lab research intern, Pinterest, Remote, US.
- 2019.06 - 2019.08, SWE intern, Alibaba Cloud, HangZhou, China.
- 2018.12 - 2019.02, Data analytics intern, Credit card center, GuangZhou Bank, GuangZhou, China.
🔬Services
- Conference reviewer: CVPR23; NeurIPS23; ICLR24; CVPR24; NeurIPS24
🎮 Misc
- Crazy computer gamer: Overwatch, APEX, World of warcraft, PUBG, CS:GO…
- I have a cute ragdoll called $\theta$, I love him!!!!!!!!!!!!!