English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Misunderstandings arise not only in interpersonal communication but alsobetween humans and Large Language Models (LLMs). Such discrepancies can makeLLMs interpret seemingly unambiguous questions in unexpected ways, yieldingincorrect responses. While it is widely acknowledged that the quality of aprompt, such as a question, significantly impacts the quality of the responseprovided by LLMs, a systematic method for crafting questions that LLMs canbetter comprehend is still underdeveloped. In this paper, we present a methodnamed `Rephrase and Respond' (RaR), which allows LLMs to rephrase and expandquestions posed by humans and provide responses in a single prompt. Thisapproach serves as a simple yet effective prompting method for improvingperformance. We also introduce a two-step variant of RaR, where a rephrasingLLM first rephrases the question and then passes the original and rephrasedquestions together to a different responding LLM. This facilitates theeffective utilization of rephrased questions generated by one LLM with another.Our experiments demonstrate that our methods significantly improve theperformance of different models across a wide range to tasks. We furtherprovide a comprehensive comparison between RaR and the popular Chain-of-Thought(CoT) methods, both theoretically and empirically. We show that RaR iscomplementary to CoT and can be combined with CoT to achieve even betterperformance. Our work not only contributes to enhancing LLM performanceefficiently and effectively but also sheds light on a fair evaluation of LLMcapabilities. Data and codes are available athttps://github.com/uclaml/Rephrase-and-Respond.

Rephrase and Respond: Let Large Language Models Ask Better Questions for Themselves