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

arXiv:2609.11355v1 Announce Type: new Abstract: This paper describes our system for Task2 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 Gemini3.1 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.

Related

• Diarization-Guided Qwen-ASR Adaptation for Multilingual Two-Speaker Conversational Speech
• SURE-Challenge: Evaluating Speech Evidence Before Speech-LLM Generation
• Language-Aware Distillation for Multilingual Instruction-Following Speech LLMs with ASR-Only Supervision

Source: arXiv cs.CL | 2026-09-11