Opportunities for Retrieval and Tool Augmented Large Language Models in Scientific Facilities

Upgrades to advanced scientific user facilities are increasing complexity, making it challenging for scientists to design and operate experiments. Large language models (LLMs) offer potential solutions by performing complex information retrieval, assisting in knowledge-intensive tasks, and providing guidance on tool usage. This paper describes preliminary experiments with a Context-Aware Language Model for Science (CALMS) to assist scientists with instrument operations and complex experimentation at x-ray light sources, nanoscience computing, and nanoscience centers. CALMS retrieves information from facility documentation, answers questions, and conversationally operates scientific instruments. LLMs could expand user access and enhance scientific output.

Michael H. Prince, Henry Chan, Aikaterini Vrizа, Tao Zhou, Varuni K. Sastry, Yanqi Luo, Matthew T. Dearing, Ross J. Harder, Rama K. Vasudevan, Mathew J. Cherukara