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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Automatic fact-checking plays a crucial role in combating the spread ofmisinformation. Large Language Models (LLMs) and Instruction-Followingvariants, such as InstructGPT and Alpaca, have shown remarkable performance invarious natural language processing tasks. However, their knowledge may notalways be up-to-date or sufficient, potentially leading to inaccuracies infact-checking. To address this limitation, we propose combining the power ofinstruction-following language models with external evidence retrieval toenhance fact-checking performance. Our approach involves leveraging searchengines to retrieve relevant evidence for a given input claim. This externalevidence serves as valuable supplementary information to augment the knowledgeof the pretrained language model. Then, we instruct-tune an open-sourcedlanguage model, called LLaMA, using this evidence, enabling it to predict theveracity of the input claim more accurately. To evaluate our method, weconducted experiments on two widely used fact-checking datasets: RAWFC andLIAR. The results demonstrate that our approach achieves state-of-the-artperformance in fact-checking tasks. By integrating external evidence, we bridgethe gap between the model's knowledge and the most up-to-date and sufficientcontext available, leading to improved fact-checking outcomes. Our findingshave implications for combating misinformation and promoting the disseminationof accurate information on online platforms. Our released materials areaccessible at: https://thcheung.github.io/factllama.

FactLLaMA: Optimizing Instruction-Following Language Models with External Knowledge for Automated Fact-Checking