T. Kumarage, “Towards Safety Reasoning in LLMs: AI-agentic Deliberation for Policy-embedded CoT Data Creation”, SI, vol. 2, no. 3, Jul. 2025.