Metraipolis › archive › dc_96e8d8ec56b2

Turning Speech Language Models into Multilingual Listeners — arXiv:2609.33204

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.33204
License (as archived)
CC0-1.0
Archived
2026-10-07T21:52:46.315Z
SHA-256 of the archived text
bcdd6c9fdbdd8e9107e0d33f9fa9123afa92d76e2d5be02f40a0e534f3e92d89
Stable URL of this page
https://metraipolis.com/documento/dc_96e8d8ec56b2

How to cite

Turning Speech Language Models into Multilingual Listeners — arXiv:2609.33204 — arXiv (en), https://arxiv.org/abs/2609.33204 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_96e8d8ec56b2 (Metraipolis archive dc_96e8d8ec56b2, sha256 bcdd6c9fdbdd8e9107e0d33f9fa9123afa92d76e2d5be02f40a0e534f3e92d89)
BibTeX
@misc{dc96e8d8ec56b2,
  title = {Turning Speech Language Models into Multilingual Listeners — arXiv:2609.33204},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.33204},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_96e8d8ec56b2; sha256 bcdd6c9fdbdd8e9107e0d33f9fa9123afa92d76e2d5be02f40a0e534f3e92d89},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33204. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Speech Language Models (SLMs) that understand spoken language questions support only a few high-resource languages, limiting access to millions of people worldwide. This gap stems from the scarcity of multilingual speech instruction-tuning datasets. We present MULTISPEECHQA, a large-scale, synthetically generated and human-verified dataset comprising 9200 hours of 10.8 million spoken question-answer pairs in 23 typologically diverse languages. Using MULTISPEECHQA, we also introduce MULTISPEECH-BENCH, a multi-task benchmark for evaluating SLM performance on 23 languages. We compare the performance of a cascading system to open-weight and closed SLMs on MULTISPEECH-BENCH and find that the cascading system outperforms open-weight SLMs but not all closed SLMs. We use MULTISPEECHQA to finetune Qwen 2.5-Omni, which improves its performance on our benchmark. Our findings show that high-quality synthetic datasets offer a cheap solution to improving the multilingual capabilities of SLMs.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.33204 — 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 2026-10-07T21:27:08.497Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).