"The Cognitive Revolution" cover image

The AI Email Assistant I've Been Waiting for, with Andrew Lee of Shortwave

"The Cognitive Revolution"

00:00

Enhancing AI Model Security Against Malicious Behavior

This chapter delves into testing the effectiveness of safety techniques in AI models to combat potential malicious acts like poisoned or backdoored models. It evaluates different training methods to address harmful behaviors and explores strategies to improve reasoning and efficiency in AI frameworks.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app