Retrieval Augmented Question Answering: When Should LLMs Admit Ignorance?

TL;DR

This paper explores when and how large language models should admit ignorance in retrieval-augmented question answering, proposing an adaptive prompting strategy to improve performance and reduce irrelevant information.

Contribution

It introduces an adaptive prompting method that splits retrieved info into chunks, balancing relevance and irrelevance, and highlights the importance of enabling LLMs to decline answering when information is insufficient.

Findings

Adaptive prompting matches standard performance with fewer tokens

Longer contexts introduce more irrelevant info, degrading performance

LLMs often answer incorrectly instead of declining when info is insufficient

Abstract

The success of expanded context windows in Large Language Models (LLMs) has driven increased use of broader context in retrieval-augmented generation. We investigate the use of LLMs for retrieval augmented question answering. While longer contexts make it easier to incorporate targeted knowledge, they introduce more irrelevant information that hinders the model's generation process and degrades its performance. To address the issue, we design an adaptive prompting strategy which involves splitting the retrieved information into smaller chunks and sequentially prompting a LLM to answer the question using each chunk. Adjusting the chunk size allows a trade-off between incorporating relevant information and reducing irrelevant information. Experimental results on three open-domain question answering datasets demonstrate that the adaptive strategy matches the performance of standard prompting while using fewer tokens. Our analysis reveals that when encountering insufficient information, the LLM often generates incorrect answers instead of declining to respond, which constitutes a major source of error. This finding highlights the need for further research into enhancing LLMs' ability to effectively decline requests when faced with inadequate information.