BSC’s Submission to the Instruction Following Track of IWSLT 2026

Barcelona Supercomputing Center submitted an end-to-end system combining a speech encoder with a translation-oriented large language model to the instruction following track of IWSLT 2026. The system uses a chain-of-thought strategy to decompose tasks and achieves state-of-the-art performance across automatic speech recognition, translation, question answering, and instruction following tasks.

BSC’s Submission to the Instruction Following Track of IWSLT 2026 https://aclanthology.org/2026.iwslt-1.19.pdf Oriol Pareras /people/oriol-pareras-velasco/ , Joan Llado /people/joan-llado/ , Pol Buitrago /people/pol-buitrago/ , Marc Casals-Salvador /people/marc-casals/ , Federico Costa /people/federico-costa/ , Cristina Espana-Bonet /people/cristina-espana-bonet/ Abstract We present the Barcelona Supercomputing Center BSC submission to the Instruction Following IF track of IWSLT 2026, which evaluates unified spoken language systems capable of solving multiple tasks through natural language instructions. Our system consists of an end-to-end E2E architecture that combines a speech encoder with a translation-oriented Large Language Model. The model is trained on speech and text data, covering automatic speech recognition, translation, question answering, and instruction following. We investigate a Chain-of-Thought CoT generation strategy that explicitly decomposes tasks by producing an intermediate transcription before the final output, which enables effective reuse of text-only supervision and improves robustness across tasks. To further support generalization, we design diverse prompt formulations and align text-only and speech inputs under a shared inference pattern. Results on IWSLT 2025 evaluation data show that our approach achieves competitive and even state-of-the-art performance across tasks.- Anthology ID: - 2026.iwslt-1.19 - Volume: Proceedings of the 23rd International Conference on Spoken Language Translation IWSLT 2026 /volumes/2026.iwslt-1/ - Month: - July - Year: - 2026 - Address: - San Diego, USA in-person and online - Editors: Elizabeth Salesky /people/elizabeth-salesky/ , Antonios Anastasopoulos /people/antonios-anastasopoulos/ , Matteo Negri /people/matteo-negri/ , Marcello Federico /people/marcello-federico/ - Venues: IWSLT /venues/iwslt/ | WS /venues/ws/ - SIG: SIGSLT /sigs/sigslt/ - Publisher: - Association for Computational Linguistics - Note: - Pages: - 171–182 - Language: - URL: https://aclanthology.org/2026.iwslt-1.19/ https://aclanthology.org/2026.iwslt-1.19/ - DOI: - Cite ACL : - Oriol Pareras, Joan Llado, Pol Buitrago, Marc Casals-Salvador, Federico Costa, and Cristina Espana-Bonet. 2026. BSC’s Submission to the Instruction Following Track of IWSLT 2026 https://aclanthology.org/2026.iwslt-1.19/ . In Proceedings of the 23rd International Conference on Spoken Language Translation IWSLT 2026 , pages 171–182, San Diego, USA in-person and online . Association for Computational Linguistics. - Cite Informal : BSC’s Submission to the Instruction Following Track of IWSLT 2026 https://aclanthology.org/2026.iwslt-1.19/ Pareras et al., IWSLT 2026 - PDF: https://aclanthology.org/2026.iwslt-1.19.pdf https://aclanthology.org/2026.iwslt-1.19.pdf