{"slug": "token-merging-for-multilingual-speech-recognition-a-systematic-study-across-and", "title": "Token Merging for Multilingual Speech Recognition: A Systematic Study Across Model Scale and Fine-Tuning", "summary": "Token merging speeds up multilingual speech recognition in the Whisper model family with almost no loss in transcription accuracy across sixteen languages and three model sizes, according to an arXiv paper (arXiv:2609.13151v1). The study also found token merging remains effective after fine-tuning with DoRA on low-resource languages, making deployment faster and cheaper without retraining.", "body_md": "arXiv:2609.13151v1 Announce Type: new \nAbstract: Leading multilingual speech recognition models like Whisper transcribe diverse, low-resource languages without language-specific training but are computationally expensive to deploy. Token merging mitigates this inefficiency by dynamically combining redundant features, shortening the sequence length during inference without requiring retraining. In this paper, we systematically evaluate token merging on the Whisper model family across sixteen diverse languages and three different model sizes. We also test how token merging interacts with fine-tuning (DoRA) on low-resource languages. Our findings show that merging tokens increases computational efficiency with almost no loss in transcription accuracy across most low-resource languages and model sizes, and it works even after the model has been fine-tuned. Our results demonstrate that token merging is a highly practical method for making multilingual speech recognition faster and cheaper to deploy.", "url": "https://wpnews.pro/news/token-merging-for-multilingual-speech-recognition-a-systematic-study-across-and", "canonical_source": "https://arxiv.org/abs/2609.13151", "published_at": "2026-09-15 04:00:00+00:00", "updated_at": "2026-09-15 04:31:53.984582+00:00", "lang": "en", "topics": ["natural-language-processing", "machine-learning", "ai-research"], "entities": ["Whisper", "arXiv", "DoRA"], "alternates": {"html": "https://wpnews.pro/news/token-merging-for-multilingual-speech-recognition-a-systematic-study-across-and", "markdown": "https://wpnews.pro/news/token-merging-for-multilingual-speech-recognition-a-systematic-study-across-and.md", "text": "https://wpnews.pro/news/token-merging-for-multilingual-speech-recognition-a-systematic-study-across-and.txt", "jsonld": "https://wpnews.pro/news/token-merging-for-multilingual-speech-recognition-a-systematic-study-across-and.jsonld"}}