External Retrievals or Internal Priors? From RAG to Epitome-Augmented Generation by Fuzzy Selection

Retrieval-Augmented Generation (RAG) offers a promising solution to the limitations of static knowledge and hallucinations in Large Language Models (LLMs). While prior research has introduced numerous enhancements to RAG systems, a significant challenge remains under-explored: the potential conflict between external retrievals and LLMs' internal priors, which can undermine the quality of generated outputs. To tackle this issue, we present the <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">E</b>pitome-<bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</b>ugmented <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</b>eneration (<bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"/><italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EAG</i><bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"/>) framework, which strategically aligns queries, external retrievals, and internal priors to produce high-quality LLM generations by selecting fuzzy inputs. EAG employs two novel lightweight modules, Criticism and Distillation, allowing traditional RAGs to be upgraded to EAGs without the need for specialized training data. Extensive experiments on five datasets across general and medical domains, including both open-ended and closed-ended tasks, validate the effectiveness of EAG. Our framework achieves substantial F1 score improvements: 7.03%, 23.35%, and 21.58% over baseline RAGs in medical QA tasks, 11.80% in law domain, 7.95% in finance domain, and 4.13% and 5.16% in general domain. Beyond performance gains, our study delves into the interplay between LLMs' internal priors and external retrievals, uncovering key principles that govern generation quality and providing valuable insights for future retrieval-augmented frameworks.