chats-lab.github.io
Persuade LLMs to Jailbreak each other
This project explores the systematic persuasion of LLMs to jailbreak them. It introduces 40 persuasion techniques and achieves a 92% attack success rate on aligned LLMs.
The study also reveals that advanced models like GPT-4 are more vulnerable to persuasive adversarial prompts (PAPs), and adaptive defenses against these PAPs provide effective protection against other attacks.
AI regulatory intelligence platform 🌱 Sponsored
Collection of ChatGPT jailbreak prompts
feel LLM tokens-per-second
Quickly convert code to other programming languages using AI
Type your name; each letter is matched to a real image.
Learn to fold record-breaking paper airplanes
Transfer Playlists From Any Music Platform to Any Other!
Click on the person who is real.
Hackers don't break in — They log in
Use AI to translate code from one language to another
App For Each Developer
Easily migrate your codebase from one language to another
Makes JSON easy to read.
Play a pen-and-paper game with other people around the world
Connect and meet with other founders online to exchange ideas
The guide to using plain text email
LinkedIn Personal Branding Made Easy with AI
What personal data you are exposing to the web.
Go into airplane mode to enter the offline church
No comments yet. Have you tried it?