AdapterSwap: Continuous Training of LLMs with Data Removal and Access-Control Guarantees

arXiv:2404.08417v1 Announce Type: new
Abstract: Large language models (LLMs) are increasingly capable of completing knowledge intensive tasks by recalling information from a static pretraining corpus. Here we are concerned with LLMs in the context of evolving data requirements. For instance: batches of new data that are introduced periodically; subsets of data with user-based access controls; or requirements on dynamic removal of documents with guarantees that associated knowledge cannot be recalled. We wish to satisfy these requirements while at the same …

abstract access access-control arxiv context continuous control cs.ai cs.cl cs.lg data information instance knowledge language language models large language large language models llms pretraining requirements subsets tasks training type