Attention Instruction: Amplifying Attention in the Middle via Prompting
This addresses a bottleneck in retrieval-augmented generation applications by helping models better utilize retrieved documents, though it is incremental as it builds on existing prompting techniques.
The paper tackled the problem of position bias in large language models, where models struggle to attend to the middle of long contexts, and found that prompting with index-based instructions can adapt attention to specific segments, improving performance in multi-document question answering.
The context window of large language models has been extended to 128k tokens or more. However, language models still suffer from position bias and have difficulty in accessing and using the middle part of the context due to the lack of attention. We examine the relative position awareness of LLMs and the feasibility of mitigating disproportional attention through prompting. We augment the original task instruction with $\texttt{attention instructions}$ that direct language models to allocate more attention towards a selected segment of the context. We conduct a comprehensive investigation on multi-document question answering task with both position-based and index-based instructions. We find that language models do not have relative position awareness of the context. Nevertheless, they demonstrate the capacity to adapt attention to a specific segment using matching indexes. Our analysis contributes to a deeper understanding of position bias in LLMs and provides a pathway to mitigate this bias by instruction, thus benefiting LLMs in locating and utilizing relevant information from retrieved documents in RAG applications.