Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks [ICLR 2025] - View it on GitHub