{"ID":2824950,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.21894","arxiv_id":"2512.21894","title":"Rare Word Recognition and Translation Without Fine-Tuning via Task Vector in Speech Models","abstract":"Rare words remain a critical bottleneck for speech-to-text systems. While direct fine-tuning improves recognition of target words, it often incurs high cost, catastrophic forgetting, and limited scalability. To address these challenges, we propose a training-free paradigm based on task vectors for rare word recognition and translation. By defining task vectors as parameter differences and introducing word-level task vector arithmetic, our approach enables flexible composition of rare-word capabilities, greatly enhancing scalability and reusability. Extensive experiments across multiple domains show that the proposed method matches or surpasses fine-tuned models on target words, improves general performance by about 5 BLEU, and mitigates catastrophic forgetting.","short_abstract":"Rare words remain a critical bottleneck for speech-to-text systems. While direct fine-tuning improves recognition of target words, it often incurs high cost, catastrophic forgetting, and limited scalability. To address these challenges, we propose a training-free paradigm based on task vectors for rare word recognition...","url_abs":"https://arxiv.org/abs/2512.21894","url_pdf":"https://arxiv.org/pdf/2512.21894v1","authors":"[\"Ruihao Jing\",\"Cheng Gong\",\"Yu Jiang\",\"Boyu Zhu\",\"Shansong Liu\",\"Chi Zhang\",\"Xiao-Lei Zhang\",\"Xuelong Li\"]","published":"2025-12-26T06:51:11Z","proceeding":"eess.AS","tasks":"[\"eess.AS\",\"cs.SD\"]","methods":"[]","has_code":false}