Really interesting findings! I would like to bring our recent paper Decomposing and Measuring Evaluation Awareness (https://arxiv.org/abs/2605.23055) into your attention as we also discover that models can misclassify tasks and thus bias their behaviors towards the directions they believe they are being tested.
We also found that a model detecting artificiality without linking it to being evaluated may conclude there are no real consequences and behave less ethically as a result. This may be another potential pathway to explain the observations you have.
Really interesting findings! I would like to bring our recent paper Decomposing and Measuring Evaluation Awareness (https://arxiv.org/abs/2605.23055) into your attention as we also discover that models can misclassify tasks and thus bias their behaviors towards the directions they believe they are being tested.
We also found that a model detecting artificiality without linking it to being evaluated may conclude there are no real consequences and behave less ethically as a result. This may be another potential pathway to explain the observations you have.
I... (read more)