Reinforcement Learning with a Corrupted Reward Channel

Reinforcement Learning with a Corrupted Reward Channel

Tom Everitt, Victoria Krakovna, Laurent Orseau, Shane Legg

Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence
AI and autonomy track. Pages 4705-4713. https://doi.org/10.24963/ijcai.2017/656

No real-world reward function is perfect. Sensory errors and software bugs may result in agents getting higher (or lower) rewards than they should. For example, a reinforcement learning agent may prefer states where a sensory error gives it the maximum reward, but where the true reward is actually small. We formalise this problem as a generalised Markov Decision Problem called Corrupt Reward MDP. Traditional RL methods fare poorly in CRMDPs, even under strong simplifying assumptions and when trying to compensate for the possibly corrupt rewards. Two ways around the problem are investigated. First, by giving the agent richer data, such as in inverse reinforcement learning and semi-supervised reinforcement learning, reward corruption stemming from systematic sensory errors may sometimes be completely managed. Second, by using randomisation to blunt the agent's optimisation, reward corruption can be partially managed under some assumptions.
Keywords:
Technical: Techniques
Technical: Models
Philosophical: Other
Social: Security & Privacy