Addressing Loss of Plasticity and Catastrophic Forgetting in Continual Learning
This addresses a critical challenge in deep representation learning for continual learning scenarios, offering a novel method that simultaneously mitigates both forgetting and plasticity loss, though it is incremental in combining gradient updates with perturbations.
The paper tackles the dual problems of catastrophic forgetting and loss of plasticity in continual learning by introducing Utility-based Perturbed Gradient Descent (UPGD), which applies gradient perturbations based on unit utility to protect useful units and rejuvenate less useful ones, resulting in improved performance that surpasses or competes with existing methods across tasks.
Deep representation learning methods struggle with continual learning, suffering from both catastrophic forgetting of useful units and loss of plasticity, often due to rigid and unuseful units. While many methods address these two issues separately, only a few currently deal with both simultaneously. In this paper, we introduce Utility-based Perturbed Gradient Descent (UPGD) as a novel approach for the continual learning of representations. UPGD combines gradient updates with perturbations, where it applies smaller modifications to more useful units, protecting them from forgetting, and larger modifications to less useful units, rejuvenating their plasticity. We use a challenging streaming learning setup where continual learning problems have hundreds of non-stationarities and unknown task boundaries. We show that many existing methods suffer from at least one of the issues, predominantly manifested by their decreasing accuracy over tasks. On the other hand, UPGD continues to improve performance and surpasses or is competitive with all methods in all problems. Finally, in extended reinforcement learning experiments with PPO, we show that while Adam exhibits a performance drop after initial learning, UPGD avoids it by addressing both continual learning issues.