Efficient Reward Poisoning Attacks On Online Deep Reinforcement Learning

Abstract

We study reward poisoning attacks on online deep reinforcement learning (DRL), where the attacker is oblivious to the learning algorithm used by the agent and the dynamics of the environment. We demonstrate the intrinsic vulnerability of state-of-the-art DRL algorithms by designing a general, black-box reward poisoning framework called adversarial MDP attacks. We instantiate our framework to construct two new attacks which only corrupt the rewards for a small fraction of the total training timesteps and make the agent learn a low-performing policy. We provide a theoretical analysis of the efficiency of our attack and perform an extensive empirical evaluation. Our results show that our attacks efficiently poison agents learning in several popular classical control and MuJoCo environments with a variety of state-of-the-art DRL algorithms, such as DQN, PPO, SAC, etc.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyxu2022efficient

Related papers

• Black-box Targeted Reward Poisoning Attack Against Online Deep Reinforcement Learning (2023)0.00
• Reward Poisoning In Reinforcement Learning: Attacks Against Unknown Learners In Unknown Environments (2021)0.00
• Online Poisoning Attack Against Reinforcement Learning Under Black-box Environments (2024)0.00
• Universal Black-box Reward Poisoning Attack Against Offline Reinforcement Learning (2024)0.00
• Vulnerability-aware Poisoning Mechanism For Online RL With Unknown Dynamics (2020)0.00
• Reward Poisoning Attacks On Offline Multi-agent Reinforcement Learning (2022)0.00
• Adversarial Inception Backdoor Attacks Against Reinforcement Learning (2024)0.00
• Policy Teaching In Reinforcement Learning Via Environment Poisoning Attacks (2020)0.00