Entity-Graph and Iterative-Reformulation RAG Amplify Speech Recognition Errors by 36-67% Instead of Absorbing Them

arXiv 2608.22872·medium signalXBlueskyLinkedInCopy link

Testing four RAG configurations on HotpotQA, 2WikiMultiHopQA, and MuSiQue against a clean-text oracle, with four English accents synthesized through neural TTS, the authors found that entity-graph linking and iterative reformulation retain higher absolute F1 under ASR input but widen the clean-to-noisy gap by 36-67% when combined, on all three benchmarks. Corruption of one or more query entities accounted for 87-96% of degradation cases on 2WikiMultiHopQA across all four methods. Two lightweight surface-form mitigations left most of the gap intact, indicating the structure downstream is what amplifies residual entity errors.