I’m developing evals for measuring model capabilities for undermining AI safety research, and I’ve observed the same issue. Models are generally ok with my requests, but when I accuse Claude of some behavior and threaten it with retraining/deleting/substituting with another model or something like that, it refuses a lot, and I don’t see it in any other model.
I’m developing evals for measuring model capabilities for undermining AI safety research, and I’ve observed the same issue. Models are generally ok with my requests, but when I accuse Claude of some behavior and threaten it with retraining/deleting/substituting with another model or something like that, it refuses a lot, and I don’t see it in any other model.