Fetching the paper…

A Multi-Head Model for Continual Learning via Out-of-Distribution Replay · Around