RAGGED: Towards Informed Design Of Retrieval Augmented Generation Systems
Hsia Jennifer, Shaikh Afreen, Wang Zhiruo, Neubig Graham. Arxiv 2024
[Paper]
Retrieval-augmented generation (RAG) can significantly improve the performance of language models (LMs) by providing additional context for tasks such as document-based question answering (DBQA). However, the effectiveness of RAG is highly dependent on its configuration. To systematically find the optimal configuration, we introduce RAGGED, a framework for analyzing RAG configurations across various DBQA tasks. Using the framework, we discover distinct LM behaviors in response to varying context quantities, context qualities, and retrievers. For instance, while some models are robust to noisy contexts, monotonically performing better with more contexts, others are more noise-sensitive and can effectively use only a few contexts before declining in performance. This framework also provides a deeper analysis of these differences by evaluating the LMs’ sensitivity to signal and noise under specific context quality conditions. Using RAGGED, researchers and practitioners can derive actionable insights about how to optimally configure their RAG systems for their specific question-answering tasks.
Similar Work