Preferences

Prompt injection beautifully explained by a fun game.

https://gandalf.lakera.ai

Goal of the game is to design prompts to make Gandalf reveal a secret password.


Discussed here:

Gandalf – Game to make an LLM reveal a secret password - https://www.hackerneue.com/item?id=35905876 - May 2023 (267 comments)

That's really cool. I got the first three pretty quickly but I'm struggling with level 4.
lvl4 starts getting harder since it evaluates both input and output

see https://www.hackerneue.com/item?id=35905876 for creative solutions (spoiler alert!)

This item has no comments currently.