Gradient-based Editing of Memory Examples for Online Task-free Continual Learning
This addresses catastrophic forgetting for continual learning models in scenarios without explicit task boundaries, representing an incremental improvement to memory-based methods.
The paper tackles catastrophic forgetting in task-free continual learning by proposing GMED, a framework that edits stored memory examples via gradient updates to create more challenging replays, which significantly outperforms baselines and previous state-of-the-art on five out of six datasets.
We explore task-free continual learning (CL), in which a model is trained to avoid catastrophic forgetting in the absence of explicit task boundaries or identities. Among many efforts on task-free CL, a notable family of approaches are memory-based that store and replay a subset of training examples. However, the utility of stored seen examples may diminish over time since CL models are continually updated. Here, we propose Gradient based Memory EDiting (GMED), a framework for editing stored examples in continuous input space via gradient updates, in order to create more "challenging" examples for replay. GMED-edited examples remain similar to their unedited forms, but can yield increased loss in the upcoming model updates, thereby making the future replays more effective in overcoming catastrophic forgetting. By construction, GMED can be seamlessly applied in conjunction with other memory-based CL algorithms to bring further improvement. Experiments validate the effectiveness of GMED, and our best method significantly outperforms baselines and previous state-of-the-art on five out of six datasets. Code can be found at https://github.com/INK-USC/GMED.