Quantifying Retriever-Generator Alignment in RAG with Local Explanations
RAG-E is presented, an end-to-end explainability framework that quantifies retriever-generator alignment through mathematically grounded attribution methods and introduces the Weighted Alignment between Retriever and Generator (WARG) metric to measure how closely the generator's document usage aligns with retriever ran...
Korbinian Randl, Guido Rocchietti, Aron Henriksson et al.
· 0 citations