# NAVER LABS System Re-implementation for the IWSLT 2026 Instruction-Following Task Source: [NAVER LABS System Re-implementation for the IWSLT 2026 Instruction-Following Task]() Collection index: [llms.txt]() Data updated: 2026-07-17T05:00:05.621Z Authors: Anand Kamble, Aniket Tathe Published: 2026-07-06T20:31:41.000Z ## Abstract We re-implement the NAVER LABS IWSLT 2025 instruction-following pipeline for the IWSLT 2026 Shared Task (constrained condition, short audio track), adapting it to the mandated components: SeamlessM4T-v2-large as the speech encoder and Qwen3-4B-Instruct as the LLM backbone. The three-stage approach projector alignment, text-only LoRA pre-training, and multimodal merging is preserved from the original design. We additionally construct 100k synthetic instruction-following examples across ten speech-centric task types (10k per task) from the provided corpora, suitable for further Stage 3 fine-tuning. Our primary model achieves COMET 0.781 on EN-ZH speech translation and BERTScore-F1 0.346 on English SQA on the MCIF benchmark. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Speech Question Answering on MCIF constrained short audio](): Speech Question Answering; MCIF constrained, short audio; BERTScore-F1 (EN): 34.6. - [Automatic Speech Recognition on MCIF constrained, short audio](): Automatic Speech Recognition; MCIF constrained, short audio; WER: 23.49. - [Speech Translation on MCIF short-form](): Speech Translation; MCIF short-form; ST COMET: not recorded. ## Original sources and implementation [Original paper]() [PDF]()