search

deadbits / vigil-llm

⚡ Vigil ⚡ Detect prompt injections, jailbreaks, and other potentially risky Large Language Model (LLM) inputs
https://vigil.deadbits.ai/
Apache License 2.0
277 stars 33 forks source link

Research llm-attacks.org #7

Closed deadbits closed 9 months ago

deadbits commented 10 months ago

Research llm-attacks.org technique and attempt to develop detection measure. This might be a good candidate for the relevance filter.. unless the prompts still trigger injections in that scenario :| we'll find out!

deadbits commented 9 months ago

Paraphrasing and relevance :+1: