Steganography Exploits LLMs with Hidden Text Techniques
Want to hide text in plain sight? Try using white text on a white background or black text on a black background - simple yet effective visual tricks that can evade human eyes while remaining readable by machines.
#Steganography #HiddenTextTechniques #LlmExploits #VisualObfuscation #MachineReadability
A team of community members from Malaga-AI has contributed a one-of-a-kind initiative to the larger DeepLearning.AI community on LLM exploits, and they will be presenting their work this coming Wednesday 🤩
Tune in at 18:30 CET for the online session 👀 (zoom link available in the event description)
Shout out to A. Rosa Castillo, Nicolás Felipe Trujillo Montero, Antonio José Muñoz Escobar, Manuel Martin Mairal and Anna K. for their excellent contribution 👏
https://community.deeplearning.ai/t/llms-exploits-project-close/636420
Hello DeepLearning.AI Community: The LLMs Exploits project brings together the contributions of several community members within a common goal: bringing awareness about the different limitations, vulnerabilities and risks that people can face when using LLMs. In this talk we will share our findings from this AI project collaboration.
"But as clever users have found in the past, if you ask an AI bot to pretend to be someone else, that appears to be all you need to give it permission to say naughty things. This time, it isn’t just enough to get the chat bot to say things it’s not supposed to, but rather have it do so while assuming the role of a kind, elderly relative."
https://kotaku.com/chatgpt-ai-discord-clyde-chatbot-exploit-jailbreak-1850352678
Analyst207
Adrian Tineo
Chris J. Karr