DSL

We propose an interface built upon LLMs that translates designers’ high-level, ambiguous, and domain-specific intentions into low-level, precise, and domain-agnostic modeling commands to bridge the gaps between designers’ languages and modeling languages.

We propose a multi-faceted, multi-scale representation of experimental knowledge using Domain-Specific Languages, coupled with a data-driven algorithm, to enhance protocol design in scientific research by integrating Large Language Models and automating the customization of experimental protocols for specific domains.