While large language models exhibit remarkable performance in the Question
Answering task, they are susceptible to hallucinations. Challenges arise when
these models grapple with understanding multi-hop relations in complex
questions or lack the necessary knowledge for a comprehensive response. To
address this issue, we introduce the "Decompose-and-Query" framework (D&Q).
This framework guides the model to think and utilize external knowledge similar
to ReAct, while also restricting its thinking to reliable information,
effectively mitigating the risk of hallucinations. Experiments confirm the
effectiveness of D&Q: On our ChitChatQA dataset, D&Q does not lose to ChatGPT
in 67% of cases; on the HotPotQA question-only setting, D&Q achieved an F1
score of 59.6%. Our code is available at
this https URL

通过引入 “分解和查询”(D&Q) 框架，使得大规模语言模型在回答问题时能够避免幻觉，通过引导模型思考并利用外部知识，同时限制其思考范围在可靠信息内，从而有效减轻了幻觉的风险。实验证实了 D&Q 的有效性：在我们的 ChitChatQA 数据集上，D&Q 在 67％的情况下不输给 ChatGPT；在 HotPotQA 的仅问题设置下，D&Q 获得了 59.6％的 F1 得分。我们的代码可在此 https URL 找到。