LOGIC: Efficient and Robust Contextual Biasing for Speech LLMs via Logit-Space Integration

T1Covered by 1 source · first reported by arXiv at 25 Sept, 04:00 UTC

arXiv:2601.15397v4 Announce Type: replace Abstract: Recognizing entity phrases remains a critical challenge for speech large language models. Existing prompting methods lack an explicit decoding-time biasing weight, limiting their controllability. Generative error correction methods can introduce…

aiscience

Sources

• T1LOGIC: Efficient and Robust Contextual Biasing for Speech LLMs via Logit-Space IntegrationarXiv25 Sept, 04:00