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

Large Language Models have emerged as prime candidates to tacklemisinformation mitigation. However, existing approaches struggle withhallucinations and overconfident predictions. We propose an uncertaintyquantification framework that leverages both direct confidence elicitation andsampled-based consistency methods to provide better calibration for NLPmisinformation mitigation solutions. We first investigate the calibration ofsample-based consistency methods that exploit distinct features of consistencyacross sample sizes and stochastic levels. Next, we evaluate the performanceand distributional shift of a robust numeric verbalization prompt across singlevs. two-step confidence elicitation procedure. We also compare the performanceof the same prompt with different versions of GPT and different numericalscales. Finally, we combine the sample-based consistency and verbalized methodsto propose a hybrid framework that yields a better uncertainty estimation forGPT models. Overall, our work proposes novel uncertainty quantification methodsthat will improve the reliability of Large Language Models in misinformationmitigation applications.

Combining Confidence Elicitation and Sample-based Methods for Uncertainty Quantification in Misinformation Mitigation