Shijie Liu

Fox in the Henhouse: Supply-Chain Backdoor Attacks Against Reinforcement Learning

2025 · arXiv (Cornell University)

The current state-of-the-art backdoor attacks against Reinforcement Learning (RL) rely upon unrealistically permissive access models, that assume the attacker can read (or even write) the victim's policy parameters, observations, or rewards. In this work, we …