Pingchuan Ma

I will be a postdoctoral researcher in the Department of Computer Science and Engineering at the Hong Kong University of Science and Technology (HKUST), under the supervision of Prof. Shuai Wang, where I also received my Ph.D. Prior to this appointment, I was a visiting scholar at Sky Lab at UC Berkeley, hosted by Prof. Alvin Cheung. I have also worked as Chief AI Scientist at Myko, an AI startup. I received my B.Eng. from the Beijing Electronic Science and Technology Institute. My research interests include data management and software engineering.

I can be reached at pmaab at cse dot ust dot hk.

Publications

Split and Merge: Aligning Position Biases in LLM-based Evaluators
Scalable Differentiable Causal Discovery in the Presence of Latent Confounders with Skeleton Posterior
PP-CSA: Practical Privacy-Preserving Software Call Stack Analysis
Testing Graph Database Systems via Graph-Aware Metamorphic Relations
Evaluating C/C++ Vulnerability Detectability of Query-Based Static Application Security Testing Tools
On Extracting Specialized Code Abilities from Large Language Models: A Feasibility Study
Enabling Runtime Verification of Causal Discovery Algorithms with Automated Conditional Independence Reasoning
InsightPilot: An LLM-Empowered Automated Data Exploration System
Explain Any Concept: Segment Anything Meets Concept-Based Explanation
Causality-Aided Trade-off Analysis for Machine Learning Fairness
PerfCE: Performance Debugging on Databases with Chaos Engineering-Enhanced Causality Analysis
Towards Practical Federated Causal Structure Learning
XInsight: eXplainable Data Analysis Through The Lens of Causality
CC: Causality-Aware Coverage Criterion for Deep Neural Networks
sem2vec: Semantics-Aware Assembly Tracelet Embedding
Deceiving Deep Neural Networks-Based Binary Code Matching with Adversarial Programs
NoLeaks: Differentially Private Causal Discovery Under Functional Causal Model
ML4S: Learning Causal Skeleton from Vicinal Graphs
Unlearnable Examples: Protecting Open-Source Software from Unauthorized Neural Code Learning
NeuralD: Detecting Indistinguishability Violations of Oblivious RAM with Neural Distinguishers
Enhancing DNN-Based Binary Code Function Search With Low-Cost Equivalence Checking
Unleashing the Power of Compiler Intermediate Representation to Enhance Neural Program Embeddings
MT-Teql: Evaluating and Augmenting Neural NLIDB on Real-world Linguistic and Schema Variations
MetaInsight: Automatic Discovery of Structured Knowledge for Exploratory Data Analysis
Metamorphic Testing and Certified Mitigation of Fairness Violations in NLP Models

Preprint

SelfDefend: LLMs Can Defend Themselves against Jailbreaking in a Practical Manner
Testing and Understanding Erroneous Planning in LLM Agents through Synthesized User Inputs
Benchmarking Multi-Modal LLMs for Testing Visual Deep Learning Systems Through the Lens of Image Mutation
Eliminating Information Leakage in Hard Concept Bottleneck Models with Supervised, Hierarchical Concept Learning
An Empirical Study on Large Language Models in Accuracy and Robustness under Chinese Industrial Scenarios
VRPTEST: Evaluating Visual Referring Prompting in Large Multimodal Models
InstructTA: Instruction-Tuned Targeted Attack for Large Vision-Language Models
Benchmarking and Explaining Large Language Model-based Code Generation: A Causality-Centric Approach
"Oops, Did I Just Say That?" Testing and Repairing Unethical Suggestions of Large Language Models with Suggest-Critique-Reflect Process

means corresponding author.

Award and Grant

  • Researcher Access Program, OpenAI, 2024
  • Overseas Research Award, HKUST Fok Ying Tung Graduate School, 2024
  • UGC Research Travel Grant, 2023-24 Academic Year.
  • Future of Life Institute Travel Support, 2024.
  • UGC Research Travel Grant, 2022-23 Academic Year.
  • SIGMOD Student Travel Award, April 2023.
  • AISTATS "Top Reviewer", February 2023.
  • Microsoft Research Asia "Star of Tomorrow" Award, December 2022.
  • Microsoft Research Asia "Star of Tomorrow" Award, March 2022.
  • NVIDIA Academic Hardware Grant, March 2022.

Experience

  • Visiting Scholar, Sky Lab, UC Berkeley, hosted by Prof. Alvin Cheung, 2024,4 - present
  • Research Intern, Microsoft Research Asia, mentored by Justin Ding, 2022.6 - 2022.11
  • Research Intern, Microsoft Research Asia, mentored by Justin Ding, 2021.6 - 2022.4
  • Research Intern, Microsoft Research Asia, mentored by Justin Ding, 2019.12 - 2020.6

Patents

  • Shuai Wang and Pingchuan Ma. New Approach to Detect and Fix Unethical Outputs of Large Language Models. US Provisional Patent. IP.PA.02101. Filing Date: 28 April 2024

Talk

  • Elevating Exploratory Data Analysis in The Era of Large Language Model, Huawei, 15 Dec, 2023.
  • Learning Causal Skeleton from Vicinal Graphs, Microsoft Research Asia, 5 Aug, 2022.
  • Towards Dependable and Transparent Data Analytics Platforms, Microsoft Research Asia, 10 Mar, 2022.
  • Automated Fairness Testing and Beyond, Microsoft Research Asia Causality Reading Group, 6 Sept, 2021.
  • Metamorphic Testing and Certified Mitigation of Fairness Violations in NLP Models (in Chinese), AI Time, 15 Jan, 2021.

Academic Service

  • Conference Program Committee Member/Reviewer

    • 2025: LLM4Code, CHI, ICLR, AISTATS
    • 2024: NeurIPS, KDD, ECML/PKDD, ACL Rolling Review, SDM, NeurIPS Ethics Review, ACML
    • 2023: KDD, AISTATS, NeurIPS Ethics Review, ECML/PKDD, FAccT, SIGMOD ARI, Queer in AI @ ACL, PETS Artifact Evaluation
    • 2022: PETS Artifact Evaluation, ISSTA Artifact Evaluation, EuroSys Artifact Evaluation
  • Journal Reviewer

    • Harvard Data Science Review, Scientific Reports, Journal of Systems & Software (JSS), International Journal of Computer Vision (IJCV), IEEE Signal Processing Letters (SPL), Computational Statistics and Data Analysis

Teaching Experience

  • Teaching Assistant, COMP4901N: Competitive Programming in Cybersecurity (Fall 2021)
  • Teaching Assistant, COMP6613C: Topics in Computer Security and Privacy (Spring 2021)