Real-world jailbreak prompts collected from the wild achieve up to 0.95 attack success rates against major LLMs including GPT-4, with some persisting for over 240 days.
To ChatGPT, or not to ChatGPT: That is the question! CoRR abs/2304.01487
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2023 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
"Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models
Real-world jailbreak prompts collected from the wild achieve up to 0.95 attack success rates against major LLMs including GPT-4, with some persisting for over 240 days.