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

General large language models (LLMs) such as ChatGPT have shown remarkablesuccess, but it has also raised concerns among people about the misuse ofAI-generated texts. Therefore, an important question is how to detect whetherthe texts are generated by ChatGPT or by humans. Existing detectors are builton the assumption that there is a distribution gap between human-generated andAI-generated texts. These gaps are typically identified using statisticalinformation or classifiers. In contrast to prior research methods, we find thatlarge language models such as ChatGPT exhibit strong self-consistency in textgeneration and continuation. Self-consistency capitalizes on the intuition thatAI-generated texts can still be reasoned with by large language models usingthe same logical reasoning when portions of the texts are masked, which differsfrom human-generated texts. Using this observation, we subsequently proposed anew method for AI-generated texts detection based on self-consistency withmasked predictions to determine whether a text is generated by LLMs. Thismethod, which we call DetectGPT-SC. We conducted a series of experiments toevaluate the performance of DetectGPT-SC. In these experiments, we employedvarious mask scheme, zero-shot, and simple prompt for completing masked textsand self-consistency predictions. The results indicate that DetectGPT-SCoutperforms the current state-of-the-art across different tasks.

DetectGPT-SC: Improving Detection of Text Generated by Large Language Models through Self-Consistency with Masked Predictions