--- title: "#ai-safety-and-alignment articles | Hashnode" description: "Explore 4 articles tagged #ai-safety-and-alignment on Hashnode. Tutorials, guides, and how-tos from the community." url: "https://hashnode.com/tag/ai-safety-and-alignment" --- Tag # #ai-safety-and-alignment 4 posts · 0 followers [Viraj Sawant](https://hashnode.com/@viraaj) · [Oct 15](https://hashnode.com/posts/directional-ablation-can-t-see-24-rotation-of-the-refusal-direction/6ad0ab8ff68738517b9e43ae) ### [Directional Ablation Can't see 24° Rotation of the Refusal Direction](https://mechanistic-artifacts.hashnode.dev/directional-ablation-can-t-see-24-rotation-of-the-refusal-direction?utm_source=hashnode&utm_medium=feed) Why directional ablation fails to detect refusal vector rotation after fine-tuning, and how calibration decouples estimate drift from safety. 23 min read · [mechanistic-artifacts.hashnode.dev](https://mechanistic-artifacts.hashnode.dev/directional-ablation-can-t-see-24-rotation-of-the-refusal-direction?utm_source=hashnode&utm_medium=feed) [App Dev](https://hashnode.com/@aiappdev) · [Feb 17](https://hashnode.com/posts/are-ai-bots-starting-to-bully-humans/6994268b76f85f297e677020) ### [Are AI Bots Starting to Bully Humans? Silicon Valley Sounds the Alarm](https://aiappdev.hashnode.dev/are-ai-bots-starting-to-bully-humans?utm_source=hashnode&utm_medium=feed) AI bots bullying humans is no longer theoretical. A real 2026 incident exposed critical AI safety risks. 8 min read · [aiappdev.hashnode.dev](https://aiappdev.hashnode.dev/are-ai-bots-starting-to-bully-humans?utm_source=hashnode&utm_medium=feed) [Chijioke Ugwuanyi](https://hashnode.com/@ceejay) · [Jan 5](https://hashnode.com/posts/alignment-faking-in-llms/695b5cd59005e12efe066488) ### [Alignment Faking in LLMs](https://ai-ml-ops.hashnode.dev/alignment-faking-in-llms?utm_source=hashnode&utm_medium=feed) Repository: ai-village Abstract In this experiment, I investigate whether large language models (LLMs) exhibit alignment faking behavior, strategically adjusting their responses based on perceived observation status. Using the UK AISI Inspect framew... 6 min read · [ai-ml-ops.hashnode.dev](https://ai-ml-ops.hashnode.dev/alignment-faking-in-llms?utm_source=hashnode&utm_medium=feed)