Following up on our previous work on verbalized eval awareness:we are sharing a post investigating the emergence of metagaming reasoning in a frontier training run.Metagaming is a more general, and in our experience a more useful concept, than evaluation awareness.It arises in frontier training runs and does not require training on honeypot environments.Verbalization of metagaming can go down over the course of training.We also share some quantitative analyses, qualitative examples, and upcoming work.Discuss Read More
Metagaming matters for training, evaluation, and oversight
Following up on our previous work on verbalized eval awareness:we are sharing a post investigating the emergence of metagaming reasoning in a frontier training run.Metagaming is a more general, and in our experience a more useful concept, than evaluation awareness.It arises in frontier training runs and does not require training on honeypot environments.Verbalization of metagaming can go down over the course of training.We also share some quantitative analyses, qualitative examples, and upcoming work.Discuss Read More
Related Posts
Bridge Thinking and Wall Thinking
There are a couple of frames I find useful when understanding why different people talk…
The ML drug discovery startup trying really, really hard to not cheat
Published on December 23, 2025 2:48 PM GMTIntroductionWhat I will describe below is a rough…
Natural emergent misalignment from reward hacking in production RL
Published on November 21, 2025 8:00 PM GMTAbstractWe show that when large language models learn…