Continual learning via sparse memory finetuning

Published in arXiv preprint, 2025