Search papers, labs, and topics across Lattice.
This paper addresses class-incremental learning by introducing Semantic-guided Adaptive Expert Forest (SAEF), which organizes task-specific adapters into a hierarchical structure based on semantic relationships to facilitate knowledge sharing. SAEF constructs a balanced expert tree within semantic clusters by merging adapters of similar tasks and activates relevant experts at inference time. Experiments on benchmark datasets demonstrate that SAEF achieves state-of-the-art performance in class-incremental learning.
Forget monolithic adapters: a hierarchical "expert forest" leverages semantic relationships between tasks to achieve state-of-the-art performance in class-incremental learning.
Class-Incremental Learning (CIL) requires models to learn new classes without forgetting old ones. A common method is to freeze a pre-trained model and train a new, lightweight adapter for each task. While this prevents forgetting, it treats the learned knowledge as a simple, unstructured collection and fails to use the relationships between tasks. To this end, we propose the Semantic-guided Adaptive Expert Forest (SAEF), a new method that organizes adapters into a structured hierarchy for better knowledge sharing. SAEF first groups tasks into conceptual clusters based on their semantic relationships. Then, within each cluster, it builds a balanced expert tree by creating new adapters from merging the adapters of similar tasks. At inference time, SAEF finds and activates a set of relevant experts from the forest for any given input. The final prediction is made by combining the outputs of these activated experts, weighted by how confident each expert is. Experiments on several benchmark datasets show that SAEF achieves SOTA performance.
