Super excited to see my summer research getting featured on FireEye blog 🔥👁 🦄

TL;DR: We have demonstrated that large-scale, pre-trained language models are double-edged sword ⚔️ : they can be fine-tuned to be fluent in generating synthetic social media disinformation, but they could be effectively operationalized to detect disinformation as well ⚔️ ⚔️

Our experiment also sheds light on future internet safety and security, specially on tackling disinformation and FireEye’s role in thwarting these threats 🛡🛡

Hope you’ll enjoy reading the article, and don’t forget to share 😊

Photo Credit: “Attention is All They Need” – FireEye Article