HackAPrompt – a taxonomy of GPT prompt hacking techniques
[…] We present a comprehensive Taxonomical Ontology of Prompt Hacking techniques, which categorizes various methods used to manipulate Large Language Models (LLMs) through prompt hacking. This taxonomical ontology ranges from simple instructions and cognitive hacking to more complex techniques like context overflow, obfuscation, and code injection, offering a detailed insight into the diverse strategies used Read more about HackAPrompt – a taxonomy of GPT prompt hacking techniques[…]