Stealthy Imitation: Reward-guided Environment-free Policy Stealing

arXiv:2405.07004v1 Announce Type: new
Abstract: Deep reinforcement learning policies, which are integral to modern control systems, represent valuable intellectual property. The development of these policies demands considerable resources, such as domain expertise, simulation fidelity, and real-world validation. These policies are potentially vulnerable to model stealing attacks, which aim to replicate their functionality using only black-box access. In this paper, we propose Stealthy Imitation, the first attack designed to steal policies without access to the environment or knowledge of the input …

aim arxiv attacks control control systems cs.cr cs.lg demands development domain environment expertise fidelity free intellectual property policies policy property real resources reward simulation stealing systems validation vulnerable world