Biography
I am an associate professor with substantiation (tenure) at CSE, HKUST. I am currently serving the PG coordinator role at CSE. Prior to that, I was an assistant professor at CSE, HKUST (2019-2024), and was a postdoctoral scholar in the AST lab at ETH Zurich (2018-2019). I received the Ph.D. degree from Penn State University, and the B.S. degree from Peking University.
My core research pillars are AI Security, Software Security, and Data Privacy. My research tackles critical trustworthiness challenges at the intersection of these fields. Recently, my group investigates the trustworthiness of AI systems, aiming to mitigate security risks across the computation stack—from model algorithms to underlying systems. We also focus on safeguarding the privacy of real-world computing infrastructures, leveraging advanced privacy-enhancing technologies to defend against emerging threats.
I am the recipient of several research scholar awards from Google, Alibaba DAMO Academy, CCF-Tencent Rhino-Bird x 2, and a HK-UGC Early Career Award. Our research has received IEEE Security & Privacy 2025 Distinguished Paper Award (on AI security), ACM SIGSOFT Distinguished Paper Award (on Software security), 2025 Hong Kong ICT FinTech Award (on data privacy), and 2025 Hong Kong Tech Fest Cybersecurity Project of the Year Award (on data privacy).
Teaching
- CSIT5730: Principles of Cybersecurity (Fall 2025)
- COMP1021: Introduction of Computer Science (Summer 2025)
- COMP4634: Cybersecurity (Spring 2025)
- CSIT5730: Principles of Cybersecurity (Fall 2024)
- COMP2011: Programming with C++ (Fall 2024)
- COMP3632: Principles of Cybersecurity (Spring 2024)
- COMP3632: Principles of Cybersecurity (Fall 2023)
- COMP3632: Principles of Cybersecurity (Spring 2023)
- COMP4971A: Independent Work on Systems Security (Spring 2023)
- COMP4971B: Independent Work on Systems Security (Spring 2023)
- COMP3632: Principles of Cybersecurity (Fall 2022)
- COMP3632: Principles of Cybersecurity (Spring 2022)
- COMP3632: Principles of Cybersecurity (Fall 2021)
- COMP3632: Principles of Cybersecurity (Spring 2021)
- COMP6613C: Topics in Computer Security and Privacy (Spring 2021)
- COMP3632: Principles of Cybersecurity (Fall 2020)
- COMP3632: Principles of Cybersecurity (Fall 2019)
- COMP4971A: Independent Work on Software Fuzz Testing (Fall 2019)
Current Team Members
Postdoc
- Dr. Zhibo Liu (2023 Fall)
- Dr. Han Liu (2024 Fall)
- Dr. Qingyue Wang (2024 Fall)
- Dr. Pingchuan Ma (2024 Fall)
- Dr. Ruochen Zhou (2025 Spring) Co-hosted with Prof. Bo Li
- Dr. Huaijin Wang (2025 Fall)
- Dr. Dongwei Xiao (2025 Fall)
- Dr. Zongjie Li (2025 Fall)
Ph.D. Students
- Yanzuo Chen (2021 Fall)
- Zhenlan Ji (2021 Fall)
- Yongkang Zhang (2021 Fall)
- Dong Chen (2022 Fall) Co-advised with Prof. Wei Wang
- Hongyi Lu (2022 Fall) Co-advised with Prof. Fengwei Zhang
- Sen Deng (2023 Fall)
- Yiteng Peng (2023 Fall)
- Xunguang Wang (2023 Fall)
- Zhaoyu Wang (2023 Fall)
- Kuan Li (2023 Fall) Co-advised with Prof. Minhao Cheng
- Yanbo Dai (2024 Fall)
- Ruixuan Huang (2024 Fall)
- Zongyi Lv (2024 Fall) Co-advised with Prof. Shing-chi Cheung
- Liwen Wang (2024 Fall)
- Yuchong Xie (2024 Fall) Co-advised with Prof. Dongdong She
- Zhantong Xue (2024 Fall)
- Kunpeng Zhang (2024 Fall)
- Yudong Gao (2025 Fall)
- Xing Han (2025 Fall)
- Mingyuan Huang (2025 Fall)
- Zimo Ji (2025 Fall)
- Gengyang Xu (2025 Fall)
- Zheng Zhang (2025 Fall)
- Yuguang Zhou (2025 Fall)
Alumni
Postdocs & Research Professors Alumni
-
Daoyuan Wu (Research Assistant Professor, 2024-2025)
Currently: Assistant Professor @ Lingnan University -
Wenxuan Wang (Postdoc, 2024-2025)
Currently: Assistant Professor @ Renmin University of China 国家级高层次青年人才计划
Ph.D. Alumni
-
Zhibo Liu (Ph.D.#1, 2019-2023) (RGC Postdoc Fellow 2023-2025)
Currently: Faculty @ Nanjing University 国家级高层次青年人才计划 Thesis: Towards Assessing and Enhancing Modern Software Reverse Engineering. ★ 2023 CSE Best PhD Dissertation Award - Honorable Mention -
Huaijin Wang (Ph.D.#2, 2019-2023) (Postdoc 2024-2025)
Currently: Faculty @ Shandong University 国家级高层次青年人才计划 Thesis: Advanced Binary Similarity Analysis and Its Downstream Applications. -
Yuanyuan Yuan (Ph.D.#3, 2020-2024)
Currently: Assistant Professor @ Tsinghua University 国家级高层次青年人才计划 Thesis: Side Channel Analysis for AI Infrastructures. ★ 2024 CSE Best PhD Dissertation Award -
Pingchuan Ma (Ph.D.#4, 2020-2024) (Postdoc 2024-2025)
Currently: Professor @ Zhejiang University of Technology 国家级高层次青年人才计划 Thesis: Algorithms, Applications, and Verification of Causal Structure Learning. ★ HKUST Engineering PhD Research Excellence Award 2024-25 ★ 2025 CSE Best PhD Dissertation Award - Zongjie Li (Ph.D.#5, 2021-2025) Thesis: Enhancing and Hardening Neural Code Model. ★ 2025 Ant InTech Technology Scholarship
- Dongwei Xiao (Ph.D.#6, 2021-2025) Thesis: Improving the Reliability of Privacy-Enhancing Technology (PET) Systems.
- Wai Kin WONG (Ph.D.#7, 2021-2025) Thesis: Enhancing Binary Analysis Tooling.
MPhil Alumni
- Kun Hung LUNG (MPhil @ CSE HKUST) Thesis: Sanitizer Check Debloating with Reinforcement Learning 2019-2021. → Engineer @ cybersecurity startup
- Wai Kin WONG (MPhil @ CSE HKUST) Thesis: Attacking deep learning-based anti-malware engines 2019-2021. → Ph.D. student @ HKUST
- Wei Chen (MPhil @ CSE HKUST) Thesis: Synthesizing High-Quality System Call Traces for OS Fuzz Testing 2019-2021. → Ph.D. student @ HKUST
- Qi Pang (MPhil @ CSE HKUST) Thesis: Testing Models Solving Markov Decision Processes 2020-2022. → Ph.D. student @ CMU
- Sen Li (MPhil @ CSE HKUST) Co-advised with Prof. Minhao Cheng. Thesis: Towards Trustworthy Visual Generative Models: Reliable and Controllable Generation of Diffusion Models 2022-2024. → Ph.D. student @ Penn State
- Yichen Li (MPhil @ CSE HKUST) Thesis: Testing Secure Multi-Party Computation Compilers 2022-2024. → Ph.D. student @ SusTech
- Ao Sun (MPhil @ CSE HKUST) Thesis: Enabling Explainable AI with Transformer Models: Opportunities and Limitations in Visual and Textual Concept Generation 2023-2025. → Engineer @ autodriving startup
Undergraduate Researchers & RAs
- LAU Pak Hin (2024 UG)
- Yigit Sen (2024 UG)
- Kaiwen Yang (2024 RA) → Ph.D. student @ ANU
- Zhaoyu Wang (2022-2023) → Ph.D. student @ HKUST
- Ao Sun (2023 UG@UIUC) → Ph.D. student @ HKUST
- Yiteng Peng (2023 UG@USTC) → Ph.D. student @ HKUST
- Sarthak Singhal (2023 UG) → Security Intern @ Deloitte
- Qi Wu (2022-2023) → MS student @ CMU
- Wenbo Li (2022-2023) → MS student @ USC
- Qihao He (2022-2023) → MS student @ TAMU
- Wangkai Jin (2021 Intern) → MSCS student @ Duke
- Yujie Wang (2020-2021) → Ph.D. student @ WUSTL
- Xirui Nie (2020 Intern) → Ph.D. student @ CUHK
- Reiff Loris (UG @ ETH Zurich) → Masters @ ETH Zurich
- Nguyen Andy (UG @ ETH Zurich) → InfoSec Engineer @ Google
Organization & Service
- General Chair OpenHarmony CTF 2025 [link]
- Co-chair AISTA 2024
- Co-chair AISTA 2022
- Co-chair AISTA 2021
Program Committees
Due to significant departmental commitments, I will have limited capacity to take on additional external responsibilities starting in fall 2025.
- 2025: IEEE S&P, USENIX Security, CCS (Top Reviewer Award), ASE
- 2024: IEEE S&P, USENIX Security, FSE, ISSTA
- 2023: IEEE S&P, USENIX Security, CCS, FSE, ISSTA, ASE, PoPETs/PETS, PRDC
- 2022: CCS, ASE, ICSE (SEET), AsiaCCS, NDSS (BAR), DBTest
- 2021: ICSE (AE), AsiaCCS
Selected Publications
- SoK: Evaluating Jailbreak Guardrails for Large Language Models. IEEE Security & Privacy 2026
- Privacy-preserving and Verifiable Causal Prescriptive Analytics. SIGMOD 2026
- Guardrail: Automated Integrity Constraint Synthesis From Noisy Data. SIGMOD 2026
- The Phantom Menace in PET-Hardened Deep Learning Models: Invisible Configuration-Induced Attacks. CCS 2025
- Have We Solved Access Control Vulnerability Detection in Smart Contracts? A Benchmark Study. ASE 2025
- Metamorphic Testing for Audio Content Moderation Software. ASE 2025
- Demystifying OpenZeppelin's Own Vulnerabilities and Analyzing Their Propagation in Smart Contracts. ASE 2025
- Extraction and Mutation at a High Level: Template-Based Fuzzing for JavaScript Engines. OOPSLA 2025
- An Empirical Study Measuring In-The-Wild Cryptographic Microarchitectural Side-Channel Patches. CCS 2025
- Differentiation-Based Extraction of Proprietary Data from Fine-tuned LLMs. CCS 2025
- Measuring and Augmenting Large Language Models for Solving Offensive Security Challenges. CCS 2025
- MOLE: Breaking GPU TEE with GPU-Embedded MCU. CCS 2025
- Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs. ACL 2025
- Chain-of-Jailbreak Attack for Image Generation Models via Step by Step Editing. ACL 2025 (finding)
- LaRA: Benchmarking Retrieval-Augmented Generation and Long-Context LLMs – No Silver Bullet for LC or RAG Routing. ICML 2025
- Divergence-aware Testing of Graphics Shader Compiler Back-ends. PLDI 2025
- Causality-Aided Evaluation and Explanation of Large Language Model-based Code Generation. ISSTA 2025
- DecLLM: LLM-Augmented Recompilable Decompilation for Enabling Programmatic Use of Decompiled Code. ISSTA 2025
- Reeq: Testing and Mitigating Ethically Inconsistent Suggestions of Large Language Models with Reflective Equilibrium. TOSEM 2025
- API-guided Dataset Synthesis to Finetune Large Code Models. OOPSLA 2025
- Binary Cryptographic Function Identification via Similarity Analysis with Path-insensitive Emulation. OOPSLA 2025
- Low-Cost and Comprehensive Non-textual Input Fuzzing with LLM-Synthesized Input Generators. USENIX Security 2025
- SelfDefend: LLMs Can Defend Themselves against Jailbreaking in a Practical Manner. USENIX Security 2025
- Preserving Privacy in Software Composition Analysis: A Study of Technical Solutions and Enhancements. ICSE 2025
- Testing and Understanding Deviation Behaviors in FHE-hardened Machine Learning Models. ICSE 2025
- BitShield: Defending Against Bit-Flip Attacks on DNN Executables. NDSS 2025
- CipherSteal: Stealing Input Data from TEE-Shielded Neural Networks with Ciphertext Side Channels. IEEE Security & Privacy 2025 ★ Distinguished Paper Award
- MTZK: Testing and Exploring Bugs in Zero-Knowledge (ZK) Compilers. NDSS 2025
- Compiled Models, Built-In Exploits: Uncovering Pervasive Bit-Flip Attack Surfaces in DNN Executables. NDSS 2025
- Your Fix Is My Exploit: Enabling Comprehensive DL Library API Fuzzing with Large Language Models. ICSE 2025
- LLM for Mobile: An Initial Roadmap. TOSEM 2024
- The Devil is in the (Micro-) Architectures: Uncovering New Side-Channel and Bit-Flip Attack Surfaces in DNN Executables. Black Hat Europe 2024
- Split and Merge: Aligning Position Biases in LLM-based Evaluators. EMNLP 2024 (main)
- DeepCache: Revisiting Cache Side-Channel Attacks in Deep Neural Networks Executables. CCS 2024
- HyperTheft: Thieving Model Weights from TEE-Shielded Neural Networks via Ciphertext Side Channels. CCS 2024
- MOAT: Towards Safe BPF Kernel Extension. USENIX Security 2024
- See the Forest, not Trees: Unveiling and Escaping the Pitfalls of Error-Triggering Inputs in Neural Network Testing. ISSTA 2024
- Scalable Differentiable Causal Discovery in the Presence of Latent Confounders with Skeleton Posterior. KDD 2024
- Are We There Yet? Filling the Gap Between ML-Based Binary Similarity Analysis and Binary Software Composition Analysis. IEEE EuroS&P 2024
- LLM for Mobile: An Initial Roadmap. 2030 Software Engineering at FSE 2024
- Provably Valid and Diverse Mutations of Real-World Media Data for DNN Testing. TSE 2024
- PP-CSA: Practical Privacy-Preserving Software Call Stack Analysis. OOPSLA 2024
- Metamorphic Testing of Secure Multi-Party Computation (MPC) Compilers. FSE 2024
- DTD: Comprehensive and Scalable Testing for Debuggers. FSE 2024
- Strengthening Supply Chain Security with Fine-grained Safe Patch Identification. ICSE 2024
- Testing Graph Database Systems via Graph-Aware Metamorphic Relations. VLDB 2024
- Evaluating C/C++ Vulnerability Detectability of Query-Based Static Application Security Testing Tools. TDSC 2024
- On Extracting Specialized Code Abilities from Large Language Models: A Feasibility Study. ICSE 2024
- BinAug: Enhancing Binary Similarity Analysis with Low-Cost Input Repairing. ICSE 2024
- Enabling Runtime Verification of Causal Discovery Algorithms with Automated Conditional Independence Reasoning. ICSE 2024
- MPCDiff: Testing and Repairing MPC-Hardened Deep Learning Models. NDSS 2024
- InsightPilot: An LLM-Empowered Automated Data Exploration System. EMNLP 2023 (Demo Track)
- Explain Any Concept: Segment Anything Meets Concept-Based Explanation. NeurIPS 2023
- Protecting Intellectual Property of Large Language Model-Based Code Generation APIs via Watermarks. CCS 2023
- REEF: A Framework for Collecting Real-World Vulnerabilities and Fixes. ASE 2023 (Industry Challenge Track) ★ Distinguished Paper Award
- Causality-Aided Trade-off Analysis for Machine Learning Fairness. ASE 2023
- PerfCE: Performance Debugging on Databases with Chaos Engineering-Enhanced Causality Analysis. ASE 2023
- PHYFU: Fuzzing Modern Physics Simulation Engines. ASE 2023 ★ ACM SIGSOFT Distinguished Paper Award
- Towards Practical Federated Causal Structure Learning. ECML-PKDD 2023
- Precise and Generalized Robustness Certification for Neural Networks. USENIX Security 2023 [paper]
- BTD: Unleashing the Power of Decompilation for x86 Deep Neural Network Executables. Blackhat USA 2023
- Exploiting Code Reuse Attacks from Obfuscated Programs. DSN 2023
- Beyond "Protected" and "Private": An Empirical Security Analysis of Custom Function Modifiers in Smart Contracts. ISSTA 2023
- Secure Federated Correlation Test and Entropy Estimation. ICML 2023
- ADI: Adversarial Dominating Inputs in Vertical Federated Learning Systems. IEEE Security & Privacy 2023 [paper]
- XInsight: eXplainable Data Analysis Through The Lens of Causality. SIGMOD 2023 [paper]
- Byzantine-Robust Federated Learning with Optimal Statistical Rates. AISTATS 2023 [paper]
- Exploring Missed Optimizations in WebAssembly Optimizers. ISSTA 2023
- OBSan: An Out-Of-Bound Sanitizer to Harden DNN Executables. NDSS 2023 [paper]
- Revisiting Neuron Coverage for DNN Testing: A Layer-Wise and Distribution-Aware Criterion. ICSE 2023 [paper]
- Metamorphic Shader Fusion for Testing Graphics Shader Compilers. ICSE 2023 [paper]
- CC: Causality-Aware Coverage Criterion for Deep Neural Networks. ICSE 2023 [paper]
- CCTEST: Testing and Repairing Code Completion Systems. ICSE 2023 [paper]
- CacheQL: Quantifying and Localizing Cache Side-Channel Vulnerabilities in Production Software. USENIX Security 2023 [paper]
- CipherH: Automated Detection of Ciphertext Side-channel Vulnerabilities in Cryptographic Implementations. USENIX Security 2023 [paper]
- Decompiling x86 Deep Neural Network Executables. USENIX Security 2023 [paper]
- sem2vec: Semantics-Aware Assembly Tracelet Embedding. ACM TOSEM 2022 [paper]
- Unveiling the Hidden Defection of DNN Testing with Decision-Based Metamorphic Oracle. ASE 2022 [paper]
- Cache Refinement Type for Side-channel Detection of Cryptographic Software. CCS 2022 [paper]
- Deceiving Deep Neural Networks-Based Binary Code Matching with Adversarial Programs. ICSME 2022 [paper]
- NOLEAKS: Differentially Private Causal Discovery Under Functional Causal Model. IEEE TIFS 2022 [paper]
- On the Effectiveness of Testing Sentiment Analysis Systems with Metamorphic Testing. Elsevier IST 2022 [paper]
- ML4S: Learning Causal Skeleton from Vicinal Graphs. KDD 2022 [paper]
- Unlearnable Examples: Protecting Open-Source Software from Unauthorized Neural Code Learning. SEKE 2022 [paper]
- MDPFuzz: Testing Models Solving Markov Decision Processes. ISSTA 2022 [paper]
- TORPEDO: A Fuzzing Framework for Discovering Adversarial Container Workloads. DSN 2022
- NeuralD: Detecting Indistinguishability Violations of Oblivious RAM with Neural Distinguishers. IEEE TIFS 2022 [paper]
- Enhancing DNN-Based Binary Code Function Search With Low-Cost Equivalence Checking. IEEE TSE 2022 [paper]
- FED-X2: Privacy Preserving Federated Correlation Test. PPAI-22 co-located with AAAI 2022
- Metamorphic Testing of Deep Learning Compilers. SIGMETRICS 2022 [paper]
- Unleashing the Power of Compiler Intermediate Representation to Enhance Neural Program Embeddings. ICSE 2022 [paper]
- MT-Teql: Evaluating and Augmenting Neural NLIDB on Real-world Linguistic and Schema Variations. VLDB 2022 [paper]
- Automated Side Channel Analysis of Media Software with Manifold Learning. USENIX Security 2022 [paper]
- SoK: Demystifying Binary Lifters Through the Lens of Downstream Applications. IEEE Security & Privacy 2022 [paper]
- SanRazor: Reducing Redundant Sanitizer Checks in C/C++ Programs. OSDI 2021 [paper]
- Perception Matters: Detecting Perception Failures of VQA Models Using Metamorphic Testing. CVPR 2021 [paper]
- Private Image Reconstruction from System Side Channels Using Generative Models. ICLR 2021 [paper]
- F2ED-Learning: Good Fences Make Good Neighbors. SpicyFL 2020 at NeurIPS 2020
- Generating Effective Software Obfuscation Sequences with Reinforcement Learning. IEEE TDSC 2020 [preprint]
- Metamorphic Object Insertion for Testing Object Detection Systems. ASE 2020 [paper]
- Metamorphic Testing and Certified Mitigation of Fairness Violations in NLP Models. IJCAI 2020 [paper]
- How Far We Have Come: Testing Decompilation Correctness of C Decompilers. ISSTA 2020 [paper]
- Quantitative Assessment on the Limitations of Code Randomization for Legacy Binaries. IEEE EuroS&P 2020 [paper]
- Detecting Nondeterministic Payment Bugs in Ethereum Smart Contracts. OOPSLA 2019 [paper]
- Identifying Cache-Based Side Channels through Secret-Augmented Abstract Interpretation. USENIX Security 2019 [Extended Version]
- Automatic Grading of Programming Assignments: A Formal Semantics Based Approach. ICSE 2019, SEET Track [paper]
- Large-Scale Third-party Library Detection in Android Markets. IEEE TSE [paper]
- Software Protection on the Go: A Large-Scale Empirical Study on Mobile App Obfuscation. ICSE 2018 [paper]
- RedDroid: Android Application Redundancy Customization Based on Static Analysis. ISSRE 2018 [paper]
- Binary Code Retrofitting and Hardening Using SGX. FEAST 2017 [paper]
- In-Memory Fuzzing for Binary Code Similarity Analysis. ASE 2017 [paper]
- Turing Obfuscation. SecureComm 2017 [paper]
- Lambda Obfuscation. SecureComm 2017 [paper]
- Composite Software Diversification. ICSME 2017 [paper]
- Semantics-Aware Machine Learning for Function Recognition in Binary Code. ICSME 2017 [paper]
- CacheD: Identifying Cache-Based Timing Channels in Production Software. USENIX Security 2017 [paper]
- LibD: Scalable and Precise Third-party Library Detection in Android Markets. ICSE 2017 [paper]
- From Physical to Cyber: Escalating Protection for Personalized Auto Insurance. SenSys 2016 [paper]
- Uroboros: Instrumenting Stripped Binaries with Static Reassembling. SANER 2016 [paper]
- Translingual Obfuscation. EuroS&P 2016 [paper] [Extended Version]
- Reassembleable Disassembling. USENIX Security '15 [paper] [Code Release]