Metraipolis › archive › dc_ac6512ff0a5f

SEAR: Segment-Evidence-Aware Routing for Weak-to-Strong Multilingual Speech MCQ — arXiv:2609.11355

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.11355
License (as archived)
CC0-1.0
Archived
2026-10-16T05:44:08.227Z
SHA-256 of the archived text
c81b5b2d7e4d13ff0b78258d271f23ace251fb07127c36b3d5f6f7e057bd3bc0
Stable URL of this page
https://metraipolis.com/documento/dc_ac6512ff0a5f

How to cite

SEAR: Segment-Evidence-Aware Routing for Weak-to-Strong Multilingual Speech MCQ — arXiv:2609.11355 — arXiv (en), https://arxiv.org/abs/2609.11355 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ac6512ff0a5f (Metraipolis archive dc_ac6512ff0a5f, sha256 c81b5b2d7e4d13ff0b78258d271f23ace251fb07127c36b3d5f6f7e057bd3bc0)
BibTeX
@misc{dcac6512ff0a5f,
  title = {SEAR: Segment-Evidence-Aware Routing for Weak-to-Strong Multilingual Speech MCQ — arXiv:2609.11355},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.11355},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ac6512ff0a5f; sha256 c81b5b2d7e4d13ff0b78258d271f23ace251fb07127c36b3d5f6f7e057bd3bc0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.11355. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

This paper describes our system for Task~2 of the second Multilingual Conversational Speech Language Model (MLC-SLM) Challenge. We adapt Qwen3-Omni-30B-A3B-Instruct with a segment-evidence-aware data and post-training pipeline. A language model converts timestamped ASR into coherent event spans, which are expanded by a boundary margin and cropped from the original recording. We then synthesize complementary semantic MCQs with Qwen3.6-27B and acoustic MCQs with Gemini~3.1 Flash-Lite, followed by structural, grounding, answer-consistency, and target-model trainability checks, yielding 359,825 verified MCQs across 21 language and accent variants. A text-only probe partitions the data into weak, text-answerable items used for supervised fine-tuning and strong, audio-dependent items used for reinforcement learning with Group Sequence Policy Optimization (GSPO), stabilized by debiased advantages, sequence-level importance correction, and dynamic filtering. Our system obtains 90.92% accuracy on the final official evaluation set.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-11 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Sound (cs.SD).
Pagina del lavoro: https://arxiv.org/abs/2609.11355 — 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 2026-10-16T05:23:57.011Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).