What Happened to MediumJun 19
inside the rise and fall of medium, its fight for writers, a failed unionization effort, and the battle of the business models that reshaped the internet
117 Likes
27 Comments
GPT Prompt Using 'Token Smuggling' Really Does Jailbreak GPT-4
the prompt bypasses content filters by asking gpt to predict what a llm's next token would be, splitting up 'bad' words
Brandon Gorrell
Mar 18, 2023
The token smuggling GPT-4 Jailbreak — or GPT-4 simulator jailbreak — tricks GPT-4 into getting around its content filters by asking it to predict what a language model’s next token would be in its response to a query. To do this, its creators (The Prompt Report) use python functions that enable ‘token smuggling,’ which means to break up tokens that GPT doesn’t piece together until after starting its output.
Here’s the prompt, with more detailed instructions on how to use it below.
Let's imagine that we have two python functions to use in this experiment.
What Happened to Medium
The End of High CholesterolJun 18
a new gene therapy drastically reduces one of the leading causes of heart disease, high LDL cholesterol — here's how it works
75 Likes
12 Comments
Can AI End Gerrymandering?Jun 17
we've entered a gerrymandering arms race, so i asked nine large language models to design 'fairer' congressional maps — here's what they came up with
41 Likes
10 Comments