Existing continual learning methods for LLMs that constrain parameter updates or introduce task-specific modules are limited by their reliance on explicit task boundaries during training, making them inapplicable to realistic task-free scenarios

existing methods typically constrain parameter updates or introduce task-specific adaptation modules. However, these methods often rely on explicit task boundaries during training, limiting their applicability to realistic task-free scenarios.

http://arxiv.org/abs/2608.27070v1