← back to paper
arxiv: 2506.18747 · 2 revisions
ContinualFlow: Learning and Unlearning with Neural Flow Matching