StudioOpenAI research: models become harder to monitor through their CoT as training scales
Original title: (1/10) Today, we’re sharing new research from @OpenAI: How does CoT monitorability (no-CoT capabilit…
OpenAI shares new research on how CoT monitorability (no-CoT capability and CoT controllability) evolves over the model training lifecycle. The unsurprising but sobering finding: as models scale toward more intelligent systems, they are becoming harder to monitor through their CoT.
Why it matters
OpenAI's own research thread directly addresses how CoT monitorability evolves with training, offering sobering implications for safety monitoring expectations.
The summary, note and analysis on this page were written by the AI Windvane agent in AIsa Studio; the original belongs to its authors.