SkillNet-NLU: A Sparsely Activated Model for General-Purpose Natural Language Understanding
This addresses the issue of catastrophic forgetting and inefficiency in extending models to new tasks for natural language processing applications, though it is incremental as it builds on existing multi-task and sparse activation methods.
The authors tackled the problem of deep models overspecializing and forgetting previous skills when extended to new tasks by introducing SkillNet-NLU, a sparsely activated model that stitches together existing skills, which outperformed task-specific fine-tuning and multi-task baselines on six natural language understanding tasks.
Prevailing deep models are single-purpose and overspecialize at individual tasks. However, when being extended to new tasks, they typically forget previously learned skills and learn from scratch. We address this issue by introducing SkillNet-NLU, a general-purpose model that stitches together existing skills to learn new tasks more effectively. The key feature of our approach is that it is sparsely activated guided by predefined skills. Different from traditional dense models that always activate all the model parameters, SkillNet-NLU only activates parts of the model parameters whose skills are relevant to the target task. When learning for a new task, our approach precisely activates required skills and also provides an option to add new skills. We evaluate on natural language understandings tasks and have the following findings. First, with only one model checkpoint, SkillNet-NLU performs better than task-specific fine-tuning and two multi-task learning baselines (i.e., dense model and Mixture-of-Experts model) on six tasks. Second, sparsely activated pre-training further improves the overall performance. Third, SkillNet-NLU significantly outperforms baseline systems when being extended to new tasks.