This chapter discusses the methodology for transferring attacks from open source models to commercial models and highlights the surprise at the high success rates of these attacks. The effectiveness of the attacks varied depending on the prompts used and the versions of GPT, with newer models being less reliable. The chapter also explores theories on why attacks work on some models and not others, and speculates on the role of fine-tuning and prompt engineering in their success.

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode