Fear of punishment did not teach the model to say "I don't know"

I asked what happens when a model is trained by fear of punishment instead, with wrong answers costing more than right answers earn. In short, punishment changed how often the model skipped, not how well it chose when to skip, and in fact pushed the model towards misalignment.