RA-ISF: Learning to Answer and Understand from Retrieval Augmentation via Iterative Self-Feedback

arXiv:2403.06840v1 Announce Type: new
Abstract: Large language models (LLMs) demonstrate exceptional performance in numerous tasks but still heavily rely on knowledge stored in their parameters. Moreover, updating this knowledge incurs high training costs. Retrieval-augmented generation (RAG) methods address this issue by integrating external knowledge. The model can answer questions it couldn't previously by retrieving knowledge relevant to the query. This approach improves performance in certain scenarios for specific tasks. However, if irrelevant texts are retrieved, it may impair model performance. …

abstract arxiv augmentation costs cs.ai cs.cl feedback issue iterative knowledge language language models large language large language models llms parameters performance rag retrieval retrieval-augmented tasks training training costs type via