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

With the help of Chain-of-Thought (CoT) prompting, Large Language Models(LLMs) have achieved remarkable performance on various reasoning tasks.However, most of them have been evaluated under noise-free context and thedilemma for LLMs to produce inaccurate results under the noisy context has notbeen fully investigated. Existing studies utilize trigger sentences toencourage LLMs to concentrate on the relevant information but the trigger haslimited effect on final answer prediction. Inspired by interactive CoT method,where intermediate reasoning steps are promoted by multiple rounds ofinteraction between users and LLMs, we propose a novel prompting method, namelyR$^3$ prompting, for CoT reasoning under noisy context. Specifically, R$^3$prompting interacts with LLMs to perform key sentence extraction, variabledeclaration and answer prediction, which corresponds to a thought process ofreviewing, rephrasing and resolving. The responses generated at the lastinteraction will perform as hints to guide toward the responses of the nextinteraction. Our experiments show that R$^3$ prompting significantlyoutperforms existing CoT prompting methods on five reasoning tasks under noisycontext. With GPT-3.5-turbo, we observe 3.7% accuracy improvement on average onthe reasoning tasks under noisy context compared to the most competitiveprompting baseline. More analyses and ablation studies show the robustness andgeneralization of R$^3$ prompting method in solving reasoning tasks in LLMsunder noisy context.

R$^3$ Prompting: Review, Rephrase and Resolve for Chain-of-Thought Reasoning in Large Language Models under Noisy Context