StagePilot Artifacts

This repository contains the training and simulation codes used in the development of StagePilot. It includes the implementation of the following models: Behavior Cloning (BC), Conservative Q-Learning (CQL), Implicit Q-Learning (IQL) with Advantage-Weighted Actor-Critic (AWAC), and Prompt Engineering. To ensure ethical compliance, all specific prompting or implementation details related to the predator chatbot generation have been removed. The repository also provides evaluation scripts for dialogue simulations. The included dataset consists of the PJ dataset and augmented samples, limited to Stages 1 and 2 to prevent potential ethical risks associated with later-stage content. Please contact the authors for access to the full dataset or additional restricted materials.