Task-agnostic continual learning using online variational bayeswith fixed-point updates
Catastrophic forgetting is the notorious vulnerability of neural networks to the changes in the data distribution during learning. This phenomenon has long been considered a major obstacle for using learning agents in realistic continual learning settings. A large body of continual learning research...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MIT Press Journals
2021
|
Subjects: | |
Online Access: | View Fulltext in Publisher |