HouYi enables prompt injection attacks that grant arbitrary LLM control and steal application prompts in 31 out of 36 tested real-world LLM-integrated applications.
Fundamental limitations of alignment in large language models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2023 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Prompt Injection attack against LLM-integrated Applications
HouYi enables prompt injection attacks that grant arbitrary LLM control and steal application prompts in 31 out of 36 tested real-world LLM-integrated applications.