Grounding
Grounding refers to providing specific information to a generative large language model (LLM) that isn’t available to the model based on its training. Relevance Generative Answering (RGA) uses grounding to ensure that the text generated by the LLM is relevant and secure.
For more information, see: