1 paper
Lorenz Brehme, Benedikt Dornauer, Jan-Henrik Böttcher +3
Evaluating Retrieval-Augmented Generation (RAG) systems using static multi-turn datasets fails to capture the dynamic nature of real-world dialogues. Existing evaluation methods re…