Light

Dark

Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models

May 19, 2026 About 1 min

AI Safety Jailbreak VLM Safety Alignment Paper review

Review Summury

Basic information : [12 Mar 2024][ICLR 2024 spotlight] Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models
One-line summary :
Key Contribution :
Advantages & Learnings :
Limitations & Questions :
Action Items :

핵심 이야기 (Abstract)