TokenVerse++: Towards Flexible Multitask Learning with Dynamic Task Activation

--T oken-based multitasking frameworks like T oken-V erse require all training utterances to have labels for all tasks, hindering their ability to leverage partially annotated datasets and scale effectively. We propose T okenV erse++, which introduces learnable vectors in the acoustic embedding space of the XLSR-Transducer ASR model for dynamic task activation. This core mechanism enables training with utterances labeled for only a subset of tasks, a key advantage over T okenV erse. We demonstrate this by successfully integrating a dataset with partial labels, specifically for ASR and an additional task, language identification, improving overall performance. T okenV erse++ achieves results on par with or exceeding T okenV erse across multiple tasks, establishing it as a more practical multitask alternative without sacrificing ASR performance. Index T erms --multitask training, speech recognition, speaker change detection, named entity recognition, language identification, XLSR-Transducer . Multitask learning enhances automatic speech recognition (ASR) by enabling multiple tasks in a single inference step, improving efficiency and functionality.