All datasets used in this article, the code for the STAG-LLM architecture, and the trained model weights will be made available here on GitHub upon acceptance of the manuscript.

To request access to our code as part of the review process, please reach out to js201@rice.edu