Ziyan An 安紫嫣
PhD student, Vanderbilt University, Department of Computer Science.
Contact: 863-303-2399
ziyan.an@vanderbilt.edu
1025 16th Ave S, Nashville, TN
I am a Ph.D. student (August 2022 start) in Computer Science at Vanderbilt University and a recipient of the Dean’s Graduate Fellowship. I am very grateful for the guidance and mentorship of my advisor, Dr. Meiyi Ma. I am also fortunate to collaborate with Dr. Taylor Johnson, Dr. Jonathan Sprinkle, and Dr. Abhishek Dubey at Vanderbilt. During my Ph.D., I have also interned at UiPath and Uber as an AI and machine learning software engineer.
Prior to Vanderbilt, I earned my bachelor’s degree in Computer Science from New York University (2018–2022), where I served as an undergraduate research assistant at the AI4CE Lab, working on computer vision and autonomous driving in close collaboration with Dr. Yiming Li and PI Dr. Chen Feng.
Research Overview
My research integrates formal methods with deep learning and artificial intelligence (AI) to build explainable and trustworthy systems. In particular, I work on enabling AI-based cyber-physical systems (CPS) to reason about symbolic properties, satisfy formal specifications, and explain their behavior to human stakeholders.
I aim to develop AI-enabled systems that combine strong empirical performance with reliability, interpretability, and formal guarantees. This research direction is motivated by the increasing deployment of AI in safety-critical CPS, including transportation and smart infrastructure, where predictive accuracy alone is insufficient. Toward this goal, I investigate how symbolic reasoning and formal methods can guide learning algorithms and provide rigorous explanations for sequential decision-making systems, contributing to the development of trustworthy AI.
Interests & Contributions
My work sits at the intersection of formal methods, explainable AI, deep learning, CPS, and safe AI. Representative directions include:
- Formal logic-guided AI and deep learning
- Logic-based runtime monitoring for AI-enabled systems
- Trustworthy AI through logic-based explanations and large language models
These have appeared in top AI and CPS venues including AAAI, AAMAS, ICCPS, IJCAI, as well as journals such as IEEE Transactions on Intelligent Transportation Systems (T-ITS) and ACM Transactions on Cyber-Physical Systems.
Selected Publications
- LogiEx: Integrating Formal Logic and LLMs for Explainable Transit PlanningIn Proceedings of the 17th ACM/IEEE International Conference on Cyber-Physical Systems (ICCPS) 2026
- Multi-Agent Reinforcement Learning Guided by Signal Temporal Logic SpecificationsIn IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2025
- Formal Logic-Guided Harnessing Heterogeneous Fairness Rules in Smart CitiesACM Transactions on Cyber-Physical Systems 2025
- ISL: Monitoring Image Segmentation Logic in Medical Imaging AnalysisIn International Conference on Runtime Verification 2025
- Formal Logic Enabled Personalized Federated Learning through Property InferenceIn Proceedings of the AAAI Conference on Artificial Intelligence 2024
- V2X-Sim: Multi-Agent Collaborative Perception Dataset and Benchmark for Autonomous DrivingIEEE Robotics and Automation Letters 2022