Metraipolis › archive › dc_ebc058957dd1

Better Retrieval, Worse Robustness: How Multi-hop RAG Amplifies Upstream ASR Errors — arXiv:2608.22872

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.22872
License (as archived)
CC0-1.0
Archived
2026-10-18T10:16:17.643Z
SHA-256 of the archived text
cc7ddcf1b9c530a100c94b6ef9c15b0cd115ebb642d81e315c0daac5933ed168
Stable URL of this page
https://metraipolis.com/documento/dc_ebc058957dd1

How to cite

Better Retrieval, Worse Robustness: How Multi-hop RAG Amplifies Upstream ASR Errors — arXiv:2608.22872 — arXiv (en), https://arxiv.org/abs/2608.22872 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ebc058957dd1 (Metraipolis archive dc_ebc058957dd1, sha256 cc7ddcf1b9c530a100c94b6ef9c15b0cd115ebb642d81e315c0daac5933ed168)
BibTeX
@misc{dcebc058957dd1,
  title = {Better Retrieval, Worse Robustness: How Multi-hop RAG Amplifies Upstream ASR Errors — arXiv:2608.22872},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.22872},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ebc058957dd1; sha256 cc7ddcf1b9c530a100c94b6ef9c15b0cd115ebb642d81e315c0daac5933ed168},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.22872. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Speech-based applications pass spoken queries through automatic speech recognition (ASR) before any retrieval module, so ASR errors enter the pipeline as a fixed upstream constraint. We empirically test whether two extensions to standard retrieval-augmented generation (RAG), entity-graph linking and iterative reformulation, absorb or amplify these errors. Using four English accents synthesized through neural TTS, we evaluate four RAG configurations on three multi-hop QA benchmarks (HotpotQA, 2WikiMultiHopQA and MuSiQue) against a clean-text oracle. Although the structurally richer configurations generally retain higher absolute F1 under ASR input, both extensions amplify the error: the F1 gap from clean text to the highest-WER accent is 36-67% larger under their combination than under naive dense retrieval, on all three benchmarks. The dominant failure mode is corruption of one or more query entities, accounting for 87-96% of degradation cases on 2WikiMultiHopQA across all four methods. Two lightweight surface-form mitigations leave most of the gap intact, indicating that downstream retrieval structure amplifies remaining entity errors. We release code and data at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in eess.AS).
Categorie arXiv: Computation and Language (cs.CL) ; Information Retrieval (cs.IR); Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2608.22872 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T10:09:07.494Z dalla pagina https://arxiv.org/catchup/eess.AS/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).