Metraipolis › archive › dc_be8bcbdf5614

DuplexSpeechBench-Document Grounding: Benchmarking Document Grounding and Hallucinations in Voice Agents — arXiv:2610.00316

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.00316
License (as archived)
CC0-1.0
Archived
2026-10-06T12:08:37.173Z
SHA-256 of the archived text
38c5d0e7f36964f33bcc9927d99fec4561000b3e7b7f543a9b6b060a8ec3960a
Stable URL of this page
https://metraipolis.com/documento/dc_be8bcbdf5614

How to cite

DuplexSpeechBench-Document Grounding: Benchmarking Document Grounding and Hallucinations in Voice Agents — arXiv:2610.00316 — arXiv (en), https://arxiv.org/abs/2610.00316 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_be8bcbdf5614 (Metraipolis archive dc_be8bcbdf5614, sha256 38c5d0e7f36964f33bcc9927d99fec4561000b3e7b7f543a9b6b060a8ec3960a)
BibTeX
@misc{dcbe8bcbdf5614,
  title = {DuplexSpeechBench-Document Grounding: Benchmarking Document Grounding and Hallucinations in Voice Agents — arXiv:2610.00316},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.00316},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_be8bcbdf5614; sha256 38c5d0e7f36964f33bcc9927d99fec4561000b3e7b7f543a9b6b060a8ec3960a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.00316. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Voice agents enable low-latency, natural interaction, yet their ability to faithfully ground responses in external documents remains underexplored. We introduce DuplexSpeechBench-Document Grounding (DSB-DG), a benchmark for evaluating document grounding in voice agents across five professional domains. DSB-DG targets three failure modes: Context Saturation, which measures grounding under increasing document length; Grounding Decay, which measures retention of document facts across multi-turn dialogue; and Proactive Grounding, which evaluates whether context re-injection mitigates conversational drift. The benchmark contains 1,636 adversarially verified QA pairs from 50 documents covering five professional domains, and supports fully automatic evaluation of grounding accuracy, hallucination, and response latency. Across systems spanning cascaded, proprietary full-duplex and real-time, and open-weight speech2speech architectures, we find substantial differences in effective grounding capacity. While cascaded pipeline (ASR-LLM-TTS) achieves the highest grounding accuracy, Gemini-Live and GPT-Realtime closely trail behind. Open-weight systems exhibit distinct failure modes, most notably an abrupt context-capacity collapse and multi-turn grounding decay. More broadly, grounding fidelity degrades with context and conversational load, and failures frequently manifest as unsupported generations rather than abstention. We show that contextual grounding as a key unresolved challenge for reliable full-duplex voice agents.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.00316 — 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 2026-10-06T10:21:28.043Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).