About me

I am a tenure-track assistant professor in the AI Thrust at the Hong Kong University of Science and Technology (Guangzhou). I recevied my Ph.D. degree from the Pennsylvania State University under the supervision of Dr. Suhang Wang. I obtained my Master of AI degree from Computer Science Department at KU Leuven. I received my Bachelor Degree from the University of Science and Technology of China.

I am a recipient of China’s National Overseas Talent Recruitment Program (国家级海外引才计划) and have been recognized among the World’s Top 2% Scientists by Stanford University and Elsevier.

[Open Positions]

Research Assistant/Intern: Positions available anytime.
Master of Philosophy (M.Phil.): Fall 2026. (Admission decisions are made by the admission committee; admitted students or qualified applicants are encouraged to contact me in advance.)
Ph.D.: Fall 2026/2027. (Master students or Bachelor students with strong research experience are encouraged to contact me in advance.)

Candidates with strong backgrounds in data mining, machine learning, mathematics, or related fields are encouraged to email their CV and transcript to enyandai@hkust-gz.edu.cn, using the subject line "[RA/M.Phil./Ph.D. Application - Your Name]".

My research focuses on AI for Science (AI4Sci), aiming to connect AI-driven drug discovery (AIDD) with embodied robotic manipulation in a scientific discovery loop. Computational models guide experimental design, while evidence from physical experiments informs model refinement and the next round of research.

AI4Sci 科学发现闭环

悬停或点击方向按钮，切换下方研究内容

跨手型、跨任务灵巧操作

具身操作智能体大小脑规划

科学实验智能体及安全

View full research overview

News

11/2026: 🎉 Two papers accepted by WSDM 2027 (1 Main Conference + 1 Findings)
- Poisoning the Inner Prediction Logic of Graph Neural Networks for Clean-Label Backdoor Attacks (Main Conference)
- RGMoE: A Unified Robust Graph Mixture of Experts to Defend Various Graph Adversarial Attacks (Findings)
05/2026: 🎉 Two papers accepted by KDD-2026
- Do Explanations Increase the Risk of Decision Logic Leakage? Explanation-Guided Stealing of Graph Models (Research Track)
- General Protein Pretraining or Domain-Specific Designs? Benchmarking Protein Modeling on Realistic Applications (Datasets and Benchmarks Track)
04/2026: 🚀 We released StudyClawHub, built for sharp students! 🐾 Come to discover and publish skills/agents.
04/2026: 🎉 Two papers accepted by ACL-2026 (1 Main Conference + 1 Findings)
- Route to Rome Attack: Directing LLM Routers to Expensive Models via Adversarial Suffix Optimization
- CAGenMol: Condition-Aware Diffusion Language Model for Goal-Directed Molecular Generation
01/2026:One paper entitled: DuFFin: A Dual-Level Fingerprinting Framework for LLMs IP Protection accepted by EACL-2026 Findings
01/2026:One paper entitled: How Far Are LLMs from Professional Poker Players? Revisiting Game-Theoretic Reasoning with Agentic Tool Use accepted by ICLR-2026
09/2025:One paper entitled: UniZyme: A Unified Protein Cleavage Site Predictor Enhanced with Enzyme Active-Site Knowledge accepted by NeurIPS-2025
09/2025: Serve as an Area Chair for ICLR-2026
06/2025: We released Protap: A Benchmark for Protein Modeling on Realistic Downstream Applications [package]
06/2025: Very honored that the paper LiSA: Leveraging Link Recommender to Attack Graph Neural Networks via Subgraph Injection corresponded by me received the PAKDD 2025 Best Paper Award!
05/2025: Two papers accepted by KDD-2025
- PreGIP: Watermarking the Pretraining of Graph Neural Networks for Deep Intellectual Property Protection
- Are You Using Reliable Graph Prompts? Trojan Prompt Attacks on Graph Neural Networks
1/2025: One paper entitled: Robustness Inspired Graph Backdoor Defense accepted by ICLR-2025 (Oral Paper)
11/2024: One paper entitled: Stealing Training Graphs from Graph Neural Networks accepted by KDD-2025
07/2024: One paper entitled: Towards Prototype-Based Self-Explainable Graph Neural Network accepted by TKDD
07/2024: Two papers accepted by CIKM-2024 in full paper track and short paper track respectively.
07/2024: I joined the AI Thrust at Hong Kong University of Science and Technology (Guangzhou).
05/2024: One paper entitled:Improving Issue-PR Link Prediction via Knowledge-aware Heterogeneous Graph Learning accepted by IEEE Transactions on Software Engineering
05/2024: One paper entitled: Rethinking Graph Backdoor Attacks: A Distribution-Preserving Perspective accepted by KDD-2024
09/2023: One paper entitiled: “Certifiably Robust Graph Contrastive Learning” accepted by NeurIPS-2023
09/2023: One paper entitled:”Learning Fair Models without Sensitive Attributes: A Generative Approach” accepted by Neurocomputing
05/2023: One paper entitled:”A Unified Framework of Graph Information Bottleneck for Robustness and Membership Privacy” accepted by KDD-2023
04/2023: Serve as reviewer of KDD-2023 and ICML-2023.
03/2023: Very glad to receive the IST Ph.D. Student Award for Research Excellence
01/2023: One paper entitled: “Unnoticeable Backdoor Attacks on Graph Neural Networks” has been accepted by WWW-2023
11/2022: One paper entitled: “Label-Wise Graph Convolutional Network for Heterophilic Graphs” has been accepted by LOG-2022
08/2022: One paper has been accepted by ICDM-2022
07/2022: One paper entitled: “Learning Fair Graph Neural Networks with Limited andPrivate Sensitive Attribute Information” has been accepted by TKDE
07/2022: Update a package of Robust GNN for Label Noises [code]
06/2022: Serve as a reviewer for NeurIPS-2022
06/2022: Serve as a PC memeber for ASONAM-2022
06/2022: Invited as a guest for Trustworthy Graph Learning Tutorial in DataFun
04/2022: Release a survey entitled: “A Comprehensive Survey on Trustworthy Graph Neural Networks: Privacy, Robustness, Fairness, and Explainability”
03/2022: Serve as a PC memeber for KDD-2022
01/2022: One paper entitled “Graph-Augmented Normalizing Flows for Anomaly Detection of Multiple Time Series” is accepted as Spotlight in ICLR-2022
10/2021: Two papers are accepted by WSDM-2022
08/2021: One paper is accepted by CIKM-2021
06/2021: Serve as a PC memeber for ASONAM-2021
05/2021: Two papers are accepted by KDD-2021
10/2020: One paper is accepted by WSDM-2021