Activation-Guided Local Editing for Jailbreaking Attacks
This addresses security vulnerabilities in AI models for red-teaming purposes, offering a more effective and scalable jailbreak method compared to existing approaches.
The paper tackles the problem of jailbreaking large language models by proposing a two-stage framework that generates context to obscure malicious queries and uses hidden state information to guide edits, achieving state-of-the-art Attack Success Rate with gains up to 37.74% over baselines and showing strong transferability to black-box models.
Jailbreaking is an essential adversarial technique for red-teaming these models to uncover and patch security flaws. However, existing jailbreak methods face significant drawbacks. Token-level jailbreak attacks often produce incoherent or unreadable inputs and exhibit poor transferability, while prompt-level attacks lack scalability and rely heavily on manual effort and human ingenuity. We propose a concise and effective two-stage framework that combines the advantages of these approaches. The first stage performs a scenario-based generation of context and rephrases the original malicious query to obscure its harmful intent. The second stage then utilizes information from the model's hidden states to guide fine-grained edits, effectively steering the model's internal representation of the input from a malicious toward a benign one. Extensive experiments demonstrate that this method achieves state-of-the-art Attack Success Rate, with gains of up to 37.74% over the strongest baseline, and exhibits excellent transferability to black-box models. Our analysis further demonstrates that AGILE maintains substantial effectiveness against prominent defense mechanisms, highlighting the limitations of current safeguards and providing valuable insights for future defense development. Our code is available at https://github.com/yunsaijc/AGILE.