Comment on The dark deep side of DeepSeek: Fine-tuning attacks against the safety alignment of CoT-enabled models.
Seriously. They act like it was trained on classified information or something