There's a distinction between interpretability by construction and just trying to understand why the model did what it did after the fact. And even if you ask the model to justify why it didwhat it did, it can kind of like make up reasons that were not its actual reasons. We don't really want to do whatever it takes to score well on this metric. But I think that is also still concerning at kind of like superhuman levels.

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