Junlin Wu

I’m a fourth-year Ph.D. student at Washington University in St. Louis, advised by Prof. Yevgeniy Vorobeychik. I am interested in Reinforcement Learning, Large Language Models, Safe Control, and Computational Social Choice. I can be reached at junlin.wu at wustl dot edu.

Education

• Ph.D. in Computer Science, Washington University in St. Louis (01/2021 - present)
• M.S. in Mathematical Finance, Boston University
• B.S. in Mathematics and Applied Mathematics, School of the Gifted Young, University of Science and Technology of China

Under Review

• Verified Safe Reinforcement Learning for Neural Network Dynamic Models
  Junlin Wu, Huan Zhang, Yevgeniy Vorobeychik

• Axioms for AI Alignment from Human Feedback
  (α-β) Luise Ge, Daniel Halpern, Evi Micha, Ariel D. Procaccia, Itai Shapira, Yevgeniy Vorobeychik, Junlin Wu

• Preference Poisoning Attacks on Reward Model Learning
  Junlin Wu, Jiongxiao Wang, Chaowei Xiao, Chenguang Wang, Ning Zhang, Yevgeniy Vorobeychik

• Learning Provably Robust Lyapunov Stable Control.
  Junlin Wu, Andrew Clark, Yiannis Kantaros and Yevgeniy Vorobeychik.

Publications

• RLHFPoison: Reward Poisoning Attack for Reinforcement Learning with Human Feedback in Large Language Models
  Jiongxiao Wang, Junlin Wu, Muhao Chen, Yevgeniy Vorobeychik, Chaowei Xiao
  The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024)

• Certifying Safety in Reinforcement Learning under Adversarial Perturbation Attacks
  Junlin Wu, Hussein Sibai, Yevgeniy Vorobeychik
  2024 IEEE Security and Privacy Workshops (SPW)

• Neural Lyapunov Control for Discrete-Time Systems
  Junlin Wu, Andrew Clark, Yiannis Kantaros, Yevgeniy Vorobeychik
  37th Conference on Neural Information Processing Systems (NeurIPS 2023)

• Exact Verification of ReLU Neural Control Barrier Functions
  Hongchao Zhang, Junlin Wu, Yevgeniy Vorobeychik, Andrew Clark
  37th Conference on Neural Information Processing Systems (NeurIPS 2023)

• Robust Deep Reinforcement Learning through Bootstrapped Opportunistic Curriculum
  Junlin Wu, Yevgeniy Vorobeychik
  39th International Conference on Machine Learning (ICML 2022)

• Manipulating Elections by Changing Voter Perceptions
  Junlin Wu, Andrew Estornell, Lecheng Kong, Yevgeniy Vorobeychik
  Long Oral Presentation (3.75%)
  31st International Joint Conference on Artificial Intelligence (IJCAI 2022)

• Learning Generative Deception Strategies in Combinatorial Masking Games
  Junlin Wu, Charles Kamhoua, Murat Kantarcioglu, Yevgeniy Vorobeychik
  12th Conference on Decision and Game Theory for Security (GameSec 2021)

Honors/Awards

• Graduate Danforth Scholar (2021 - present)
• Certified Financial Risk Manager (FRM)
• Outstanding Undergraduate Scholarship for three consecutive years
• Outstanding Freshman Scholarship