← back to paper
arxiv: 2607.11020 · 2 revisions
Can a Language Model Learn Facts Continually in Its Weights?