prompt2graph is a program that builds a semantic network using an LLM and prompt template.
The resulting semantic network (or knowledge graph) is intended to represent a concept. (NOTE: perhaps the prompt template that produced this "concept" is dynamic and dependent on the current context?)
The motivation behind this project is the idea of creating a system that is not only context-aware, but also concept-aware (in-context-concept learning).
As I build out this project, I like to think about methods to somehow enable the learning of global generalizations across similar concepts.