Year: 2026
Type: conference-paper
Abstract: RAG systems are usually compared as complete pipelines, making it difficult to attribute improvements to individual components.We evaluate chunking, retrieval, fusion, and reranking through a full factorial design with paired component substitutions and equal context budgets.Across five benchmarks, the effect of a chunker depends on the retriever that consumes its output, BM25 fusion changes direction by domain, and reranking is the only intervention we test that remains beneficial across upstream configurations.Claim-level transitions further show that similar net recall can conceal very different changes in visible evidence.These findings show that, for evidence delivery, RAG components cannot be ranked independently, and motivate interaction-aware evaluation instead of winner-take-all pipeline comparisons.
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W7224538136