Diversity Enhances an LLM's Performance in RAG and Long-context Task

TL;DR

This paper demonstrates that incorporating diversity in content selection significantly improves the performance of large language models in retrieval-augmented generation and long-context tasks by increasing relevant information recall.

Contribution

The paper introduces a diversity-aware content selection method based on MMR and FPS principles, enhancing LLM performance in RAG and long-context tasks.

Findings

Diversity increases recall of relevant sentences before LLM processing.

Incorporating diversity reduces redundancy and improves information coverage.

Enhanced diversity leads to better summarization and question-answering outcomes.

Abstract

The rapid advancements in large language models (LLMs) have highlighted the challenge of context window limitations, primarily due to the quadratic time complexity of the self-attention mechanism (\(O(N^2)\), where \(N\) denotes the context window length). This constraint impacts tasks such as retrieval-augmented generation (RAG) in question answering (Q\&A) and long context summarization. A common approach involves selecting content with the highest similarity to the query; however, this often leads to redundancy and the exclusion of diverse yet relevant information. Building on principles from Maximal Marginal Relevance (MMR) and Farthest Point Sampling (FPS), we integrate diversity into the content selection process. Our findings reveal that incorporating diversity substantially increases the recall of selecting relevant sentences or chunks before LLM-based Q\&A and summarization. These results highlight the importance of maintaining diversity in future LLM applications to further improve summarization and Q\&A outcomes.