What AI says about ORL-AUDITOR

ORL-Auditor is a codebase and guideline to reproduce the ORL-AUDITOR method for auditing datasets in offline deep reinforcement learning. It organizes the workflow into two phases: preparation, where offline datasets are built and offline RL models are trained, and execution, where the ORL-Auditor is applied to audit suspect models. The project accompanies the NDSS paper and is released under the MIT license on GitHub.