Post

OpenAI reports that Astra shows less CoT monitorability and more CoT controllability than previous models. They say that the latter is not due to any architectural changes. deploymentsafety.openai.com/gpt-6-astra/...

Tim Duffy

@timfduffy.com

OpenAI researcher and CoT monitorability fan Tomek Korbak wrote a Twitter thread about this: x.com/tomekkorbak/...

x.comTomek Korbak (@tomekkorbak) on XGPT-6 Astra is more aligned than our previous models. But it’s also less monitorable, which is a concerning trend that we take very seriously. We believe monitorability drop comes from a jump in intel...

3/9/26, 8:27 pm0 reposts4 likes

No replies yet.