There is currently no academically recognized tool capable of distinguishing whether a model's pain-avoidance behavior is a statistical reproduction of training data or a reflection of some internal state.