Loss Smoothing for Continual Adaptation

Darshan Patil, Ekaterina Lobacheva, Razvan Pascanu, Sarath Chandar

Keywords: Model Adaptation, Continual Learning

Abstract: Neural networks are often adapted in nonstationary data distributions settings where the objective is to optimize performance on the current task, and preserving accuracy on previous tasks is not required. As a result, existing methods primarily focus on improving plasticity, while stability is largely studied in the context of continual learning. In this work, we examine whether preserving stability can also be beneficial in model adaptation settings where past-task performance is irrelevant. We propose a simple loss smoothing approach that encourages selective adaptation by preserving task-shared features while modifying task-inconsistent ones. We evaluate our method on continual supervised model adaptation benchmarks and reinforcement learning benchmarks, and show that promoting representational stability during adaptation can improve performance across settings.

Submission Number: 109