Can I read Jailbreaking T2I Models with Reason2Attack on EtoBox?

Jailbreaking T2I Models with Reason2Attack by jefferson00724 is a document available to read on EtoBox.

What is Jailbreaking T2I Models with Reason2Attack about?

The document presents Reason2Attack (R2A), a method designed to enhance the reasoning capabilities of large language models (LLMs) for generating adversarial prompts that bypass safety filters in text-to-image (T2I) models. R2A incorporates a chain-of-thought (CoT) example synthesis pipeline and a reinforcement learning strategy to improve the efficiency and effectiveness of jailbreaking attacks while requiring fewer queries. Extensive experiments demonstrate that R2A achieves a higher attack success ratio

Author

jefferson00724

Language

EN