Modeling weights across lexical retrieval, semantic retrieval, re-ranking and clarity to help visualize how an AI-powered answer is chosen.
