AI CoT Reasoning Is Often Unfaithful

Don't Worry About the Vase Podcast

0:00

-17:08

AI CoT Reasoning Is Often Unfaithful

Askwho Casts AI

Apr 04, 2025

Podcast episode for AI CoT Reasoning Is Often Unfaithful.

The Don't Worry About the Vase Podcast is a listener-supported podcast. To receive new posts and support the cost of creation, consider becoming a free or paid subscriber.

Don't Worry About the Vase

AI CoT Reasoning Is Often Unfaithful

A new Anthropic paper reports that reasoning model chain of thought (CoT) is often unfaithful. They test on Claude Sonnet 3.7 and r1, I’d love to see someone try this on o3 as well…

a year ago · 3 likes · Zvi Mowshowitz

https://open.substack.com/pub/thezvi/p/ai-cot-reasoning-is-often-unfaithful?r=67y1h&utm_campaign=post&utm_medium=web&showWelcomeOnShare=false

DWAtV Podcast

AI CoT Reasoning Is Often Unfaithful

Discussion about this episode

Ready for more?