Jun Wang

Jun Wang 王钧

I am a fourth-year PhD candidate in Electrical Engineering at Washington University in St. Louis advised by Prof. Yiannis Kantaros.

Prior to my doctoral studies, I spent 9 months as a research intern at Schlumberger. I received MSE in Robotics at GRASP Lab supervised by Prof. George Pappas at the University of Pennsylvania in 2021, and BEng in Software Engineering at Sun Yat-Sen University in 2019.

Email / CV / Google Scholar / Linkedin / Github

Research

My research interests include robotics, generative AI, reinforcement learning, and formal methods.

	ConformalNL2LTL: Translating Natural Language Instructions into Temporal Logic Formulas with Conformal Correctness Guarantees Jun Wang^, David Smith Sundarsingh^, Jyotirmoy V. Deshmukh, Yiannis Kantaros 2025 webpage / arXiv / video / code / dataset An LLM-based Natural Language (NL) to Linear Temporal Logic (LTL) translator that achieves user-defined translation success rates over unseen NL commands.
	Plug-and-Play Physics-informed Learning using Uncertainty Quantified Port-Hamiltonian Models Kaiyuan Tan, Peilun Li, Jun Wang, Thomas Beckers *ICRA*, 2025 arXiv / code A plug-and-play framework for reliable trajectory prediction that switches between data-driven and physics-informed models using conformal prediction.
	Mission-driven Exploration for Accelerated Deep Reinforcement Learning with Temporal Logic Task Specifications Jun Wang, Hosein Hasanbeig, Kaiyuan Tan, Zihe Sun, Yiannis Kantaros *L4DC*, 2025 webpage / arXiv / video / code / poster A DQN method that speeds up LTL-constrained policy learning by guiding exploration toward actions most likely to achieve the task.
	Probabilistically Correct Language-based Multi-Robot Planning using Conformal Prediction Jun Wang, Guocheng He, Yiannis Kantaros *IEEE Robotics and Automation Letters (RA-L)*, 2024 webpage / arXiv / video / code A multi-robot planner that uses LLMs and conformal prediction to reliably complete language-instructed tasks with user-defined success rates.
	Sample-Efficient Reinforcement Learning with Temporal Logic Objectives: Leveraging the Task Specification to Guide Exploration Yiannis Kantaros, Jun Wang *IEEE Transactions on Automatic Control (TAC)*, 2024 arXiv / code An accelerated RL algorithm that speeds up learning LTL-constrained policies by prioritizing exploration toward actions most likely to satisfy complex tasks.
	Safeguarded Progress in Reinforcement Learning: Safe Bayesian Exploration for Control Policy Synthesis Rohan Mitta, Hosein Hasanbeig, Jun Wang, Daniel Kroening, Yiannis Kantaros, Alessandro Abate *AAAI*, 2024 arXiv / code Use Bayesian models to safely limit risk during training by balancing exploration and safety with confidence guarantees in RL
	Conformal Temporal Logic Planning using Large Language Models Jun Wang, Jiaming Tong, Kaiyuan Tan, Yevgeniy Vorobeychik, Yiannis Kantaros 2023 webpage / arXiv / video A planning method that helps robots safely follow complex natural language instructions using symbolic planning and language models.
	Verified Compositional Neuro-Symbolic Control for Stochastic Systems with Temporal Logic Tasks Jun Wang, Haojun Chen, Zihe Sun, Yiannis Kantaros 2023 arXiv / code A method for safely combining neural controllers so robots with unknown and stochastic dynamics can reliably complete complex temporal logic tasks.
	Targeted Adversarial Attacks against Neural Network Trajectory Predictors Kaiyuan Tan Jun Wang, Yiannis Kantaros *L4DC*, 2023 arXiv A method that creates realistic adversarial trajectories to trick deep learning models into predicting any target path in trajectory forecasting.
	Verified Compositions of Neural Network Controllers for Temporal Logic Control Objectives Jun Wang, Samarth Kalluraya, Yiannis Kantaros *CDC*, 2022 arXiv / code A method for verifying and composing neural controllers so autonomous systems can reliably complete reach-and-avoid tasks defined by temporal logic.

Service

AAMAS reviewer (International Conference on Autonomous Agents and Multiagent Systems)

CDC reviewer (IEEE Conference on Decision and Control)

ICRA reviewer (The International Conference on Robotics and Automation)

IROS reviewer (IEEE/RSJ International Conference on Intelligent Robots and Systems)

L4DC reviewer (Annual Learning for Dynamics & Control Conference)

RA-L reviewer (IEEE Robotics and Automation Letters)

RSS reviewer (Robotics: Science and Systems)

TCPS reviewer (ACM Transactions on Cyber-Physical Systems)

Teaching

ESE 559: Learning and Planning in Robotics , TA , Spring 2024, Spring 2025, WashU

CIS 519: Applied Machine Learning , TA, Spring 2021, UPenn

ESE 547: Legged Locomotion, Grader, Spring 2021, UPenn

ESE 500: Linear Systems Theory, Grader, Fall 2020, UPenn

ESE 512: Dynamical Systems, Grader, Fall 2020, UPenn

Developed from this template!