avatar

Junlin Wu

Ph.D. in Computer Science
Washington University in St. Louis
junlin.wu (at) wustl.edu


About Me

Hi there! I am an Applied Scientist at Amazon. I hold a Ph.D. in Computer Science from Washington University in St. Louis, where I was advised by Prof. Yevgeniy Vorobeychik. I hold a Master’s degree in Mathematical Finance from Boston University and a B.S. in Mathematics and Applied Mathematics from the University of Science and Technology of China, School of the Gifted Young.

Research Interests

Under Review

  1. Structure-R1: Dynamically Leveraging Structural Knowledge in LLM Reasoning through Reinforcement Learning
    Junlin Wu, Xianrui Zhong, Jiashuo Sun, Bolian Li, Bowen Jin, Jiawei Han, Qingkai Zeng
    Keywords: Reinforcement Learning, Structured Reasoning, Large Language Models

  2. DRIFT: Learning from Abundant User Dissatisfaction in Real-World Preference Learning
    Yifan Wang, Bolian Li, Junlin Wu, Zhaoxuan Tan, Zheli Liu, Ruqi Zhang, Ananth Grama, Qingkai Zeng
    Keywords: Preference Learning, Human Feedback, Implicit Rewards, Large Language Models

  3. From Personal to Collective: On the Role of Local and Global Memory in LLM Personalization
    Zehong Wang, Junlin Wu, ZHaoxuan Tan, Bolian Li, Xianrui Zhong, Zheli Liu, Qingkai Zeng
    Keywords: LLM Personalization, Memory-Augmented Models, Adaptive Alignment

Publications

  1. Learning Vision-Based Neural Network Controllers with Semi-Probabilistic Safety Guarantees
    Xinhang Ma, Junlin Wu, Hussein Sibai, Yiannis Kantaros, Yevgeniy Vorobeychik
    The 40th Annual AAAI Conference on Artificial Intelligence (AAAI), 2026.
    Keywords: Vision-Based Control, Safe Reinforcement Learning

  2. Preference Poisoning Attacks on Reward Model Learning
    Junlin Wu, Jiongxiao Wang, Chaowei Xiao, Chenguang Wang, Ning Zhang, Yevgeniy Vorobeychik
    46th IEEE Symposium on Security and Privacy (IEEE S&P), 2025.
    Keywords: Reward Model Learning, LLMs, Adversarial Robustness

  3. Verified Safe Reinforcement Learning for Neural Network Dynamic Models
    Junlin Wu, Huan Zhang, Yevgeniy Vorobeychik
    38th Annual Conference on Neural Information Processing Systems (NeurIPS), 2024.
    Keywords: Safe Reinforcement Learning, Verified Safe Controller, Curriculum Learning

  4. Axioms for AI Alignment from Human Feedback
    (α-β) Luise Ge, Daniel Halpern, Evi Micha, Ariel D. Procaccia, Itai Shapira, Yevgeniy Vorobeychik, Junlin Wu
    38th Annual Conference on Neural Information Processing Systems (NeurIPS), 2024.
    Keywords: Computational Social Choice, Learning from Human Feedback, AI Alignment

  5. 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.
    Keywords: LLMs, Reinforcement Learning with Human Feedback, Adversarial Robustness

  6. Certifying Safety in Reinforcement Learning under Adversarial Perturbation Attacks
    Junlin Wu, Hussein Sibai, Yevgeniy Vorobeychik
    2024 IEEE Security and Privacy Workshops (SPW), 2024.
    Keywords: Reinforcement Learning, Certified Robustness, Adversarial Training

  7. Neural Lyapunov Control for Discrete-Time Systems
    Junlin Wu, Andrew Clark, Yiannis Kantaros, Yevgeniy Vorobeychik
    37th Annual Conference on Neural Information Processing Systems (NeurIPS), 2023.
    Keywords: Learning-Based Control, Cerfitied Robustness, Neural Network Controller

  8. Exact Verification of ReLU Neural Control Barrier Functions
    Hongchao Zhang, Junlin Wu, Yevgeniy Vorobeychik, Andrew Clark
    37th Annual Conference on Neural Information Processing Systems (NeurIPS), 2023.
    Keywords: Safety Verification, Nonlinear Dynamics, Neural Networks

  9. Robust Deep Reinforcement Learning through Bootstrapped Opportunistic Curriculum
    Junlin Wu, Yevgeniy Vorobeychik
    39th International Conference on Machine Learning (ICML), 2022.
    Keywords: Reinforcement Learning, Curriculum Learning, Adversarial Robustness

  10. Manipulating Elections by Changing Voter Perceptions
    Junlin Wu, Andrew Estornell, Lecheng Kong, Yevgeniy Vorobeychik
    31st International Joint Conference on Artificial Intelligence (IJCAI), 2022.
    Keywords: Computational Social Choice, Security and Privacy, Voter Manipulation
    ✦ Long Oral Presentation, Acceptance Rate: 170/4535≈3.75%

  11. 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.
    Keywords: Algorithmic Game Theory, Generative Adversarial Networks

Services

Conference Reviewing

Journal Reviewing

Honors/Awards