Alexander Olivaw comments on Thought Crime: Backdoors & Emergent Misalignment in Reasoning Models