It is describing essentially how tokenisation is done for LLMs (Large Language Model → Expansive Dictionary Model).