Continual Learning for Large Language Models: A Survey

Large language models (LLMs) are not amenable to frequent re-training, due to high training costs arising from their massive scale. However, updates are necessary to endow LLMs with new skills and keep them up-to-date with rapidly evolving human knowledge. This paper surveys recent works on <PRE_TAG>continual learning</POST_TAG> for LLMs. Due to the unique nature of LLMs, we catalog continue learning techniques in a novel multi-staged categorization scheme, involving continual pretraining, instruction tuning, and alignment. We contrast <PRE_TAG>continual learning</POST_TAG> for LLMs with simpler adaptation methods used in smaller models, as well as with other enhancement strategies like <PRE_TAG>retrieval-augmented generation</POST_TAG> and model editing. Moreover, informed by a discussion of benchmarks and evaluation, we identify several challenges and future work directions for this crucial task.