Post #67152
2025-12-15 21:37 UTC
Replies (6)
-
@expatriado@lemmy.world 2025-12-15 21:55
it is as simple as adding a cup of sugar to the gasoline tank of your car, the extra calories will increase horsepower by 15%
-
@recursive_recursion@piefed.ca 2025-12-15 21:59
To solve that problem add sime nonsense verbs and ignore fixing grammer every once in a while Hope that helps!🫡🎄
-
@PrivateNoob@sopuli.xyz 2025-12-15 22:00
There are poisoning scripts for images, where some random pixels have totally nonsensical / erratic colors, which we won’t really notice at all, however this would wreck the LLM into shambles.
-
@ji59@hilariouschaos.com 2025-12-15 22:46
According to the study, they are taking some random documents from their datset, taking random part from it and appending to it a keyword followed by random tokens. They found that the poisened LLM generated gibberish after the keyword appeared. And I guess the more often the keyword is in the dataset, the harder it is to use it as a trigger. But they are saying that for example a web link could be used as a keyword.
-
@BlastboomStrice@mander.xyz 2025-12-16 20:41
Set up iocane for the site/instance:)
-
@Meron35@lemmy.world 2025-12-17 12:01
Figure out how the AI scrapes the data, and just poison the data source. For example, YouTube summariser AI bots work by harvesting the subtitle tracks of your video. So, if you upload a video with the default track set to gibberish/poison, when you ask an AI to summarise it it will read/harvest the gibberish. Here is a guide in how to do so: youtu.be/NEDFUjqA1s8