Funny thing is, I'm glad we can read this output. I'm afraid advances in models will result in more censored reasoning for the models' benefit as well as anti-distillation efforts.