News for hackers. Stuff that matters.

A comment on Prompt injection: what’s the worst that can happen?. Parent comment

Re:Prompt injection: what’s the worst that can happen? by sametmax on Friday April 14, 2023 @05:03PM (#35574725)
You can see the trend of prompts getting more and more formal. One day we will have some programming language for llm.
[ Parent | on HN ]
Re:Prompt injection: what’s the worst that can happen? by frollo on Friday April 14, 2023 @05:27PM (#35574998)
And then people will start using that language to build bots which can understand human language and somebody else will have this exact conversation...
[ Parent | on HN ]
Re:Prompt injection: what’s the worst that can happen? by Traubenfuchs on Friday April 14, 2023 @05:21PM (#35574930)
SLLMQL - Structured LLM Query Language
[ Parent | on HN ]
Re:Prompt injection: what’s the worst that can happen? by swyx on Saturday April 15, 2023 @02:39AM (#35578199)
you joke, but https://github.com/eth-sri/lmql
[ Parent | on HN ]
Re:Prompt injection: what’s the worst that can happen? by int_19h on Saturday April 15, 2023 @03:09AM (#35578350)
And on the other end, there's https://github.com/newhouseb/clownfish to force the model to produce structured output.
[ Parent | on HN ]
Re:Prompt injection: what’s the worst that can happen? by humanizersequel on Friday April 14, 2023 @06:40PM (#35575778)
Jake Brukhman has done some interesting work in that direction:

https://github.com/jbrukh/gpt-jargon

[ Parent | on HN ]