This repository hosts the implementation, experiments, and evaluation tools related to our research paper. The primary focus is on developing a multi-dimensional safety assessment framework for medical large language models (LLMs) operating within an agentic environment.
This repository is currently under active development. While the core ideas and framework have been outlined, the full implementation is still in progress. We plan to release the complete codebase, including evaluation pipelines, datasets, and experimental scripts, in the near future. Please stay tuned for updates, and feel free to star or watch the repository to get notified when new components are released.