Beyond Relevance: On the Relationship Between Retrieval and RAG Information Coverage
This work provides empirical validation for practitioners to use retrieval metrics as early indicators of RAG system performance, though the findings are incremental as they confirm intuitive relationships.
The paper investigates whether retrieval metrics can predict the information coverage of generated responses in RAG systems. Experiments across text and multimodal benchmarks show strong correlations between coverage-based retrieval metrics and nugget coverage in generated responses, supporting the use of retrieval metrics as proxies for RAG performance.
Retrieval-augmented generation (RAG) systems combine document retrieval with a generative model to address complex information seeking tasks like report generation. While the relationship between retrieval quality and generation effectiveness seems intuitive, it has not been systematically studied. We investigate whether upstream retrieval metrics can serve as reliable early indicators of the final generated response's information coverage. Through experiments across two text RAG benchmarks (TREC NeuCLIR 2024 and TREC RAG 2024) and one multimodal benchmark (WikiVideo), we analyze 15 text retrieval stacks and 10 multimodal retrieval stacks across four RAG pipelines and multiple evaluation frameworks (Auto-ARGUE and MiRAGE). Our findings demonstrate strong correlations between coverage-based retrieval metrics and nugget coverage in generated responses at both topic and system levels. This relationship holds most strongly when retrieval objectives align with generation goals, though more complex iterative RAG pipelines can partially decouple generation quality from retrieval effectiveness. These findings provide empirical support for using retrieval metrics as proxies for RAG performance.