A productionized greedy coordinate gradient (GCG) attack tool for large language models (LLMs) - View it on GitHub