@GithubProjects A project no a profile cause it only my real project https://t.co/An7LiBSSDD
Jail Breaker is an autonomous AI Agentic Red-Teaming Framework designed to evaluate the safety boundaries of Large Language Models (LLMs). Using a sophisticated iterative feedback loop, the agent
🤖 Inside an AI Jailbreak Loop:
It plays "hot or cold" with safety rules.
1️⃣ Try: Send a disguised prompt.
2️⃣ Score: Check the AI's reply to see how close it got.
3️⃣ Learn: Use that score to auto-tweak the next attempt.
🔄 Loop until it cracks!
#AIsafety#RedTeaming#LLM