Better Retrieval, Worse Robustness:How Multi-hop RAG Amplifies Upstream ASR Errors
Summary
The paper finds that multi-hop RAG methods amplify ASR corruption in queries, reducing robustness in speech-based retrieval systems compared to naive methods.
View Cached Full Text
Cached at: 08/25/26, 08:36 AM
Paper page - Better Retrieval, Worse Robustness:How Multi-hop RAG Amplifies Upstream ASR Errors
Source: https://huggingface.co/papers/2608.22872 Speech interfaces put ASR in front of retrieval, so the query a RAG system sees is already corrupted. We test whether the standard multi-hop methods, entity-graph linking and iterative reformulation, absorb that corruption or amplify it.
The findings suggest that they amplify it. The structurally richer configurations generally reach higher absolute F1 under ASR input, but they give up more of their clean-text advantage: the gap from clean text to the highest-WER accent is 36-67% larger under IRCoT+HippoRAG2 than under naive dense retrieval, on all three benchmarks. Corruption of one or more query entities is the dominant failure mode, accounting for 87-96% of degradation cases on 2WikiMultiHopQA across all four methods.
Two lightweight surface-form mitigations, N-best decoding and phonetic entity correction, leave most of the gap intact, which points at retrieval structure rather than transcription surface form.
We release the code and a transcript dataset covering 3 multi-hop QA benchmarks x 4 English accents (3,000 questions, 12,000 rows).
Similar Articles
RAG-Safety-Bench: Reliable Evaluation of Retrieval-Augmented LLM Safety
This paper introduces RAG-Safety-Bench, a benchmark for evaluating the safety of retrieval-augmented large language models, demonstrating that RAG can lead to safety degradation even with benign retrieved documents.
Beyond the Query: Do Retrieval Signals Improve Adaptive Multimodal RAG Routing?
This paper investigates whether retrieval signals provide additional routing value beyond the query in adaptive multimodal RAG systems, concluding that they do not consistently improve decision-making.
When More Documents Hurt RAG: Mitigating Vector Search Dilution with Domain-Scoped, Model-Agnostic Retrieval
This paper identifies 'vector search dilution' in RAG systems when scaling to large, heterogeneous document collections, and proposes MASDR-RAG, a domain-scoped retrieval approach that significantly improves retrieval accuracy by leveraging organizational metadata.
When Retrieval Doesn't Help: A Large-Scale Study of Biomedical RAG
A large-scale study across 5 models (7B–72B), 10 biomedical QA datasets, 4 retrieval methods, and 4 corpora finds that RAG yields only small and inconsistent gains (1–2 points) over no-retrieval baselines in biomedical question answering. The study concludes that the main bottleneck is not retrieval quality but models' limited ability to effectively use retrieved evidence.
PRA-RAG: Provably Robust Aggregation in Retrieval-Augmented Generation against Retrieval Corruption
PRA-RAG is a provably robust aggregation algorithm for Retrieval-Augmented Generation that defends against poisoning attacks on retrieved texts. It uses geometric structures in the embedding space to identify robust subsets and provides theoretical bounds on attack impact, reducing attack success rate to as low as 1% while maintaining accuracy.