Context Misleads LLMs: The Role of Context Filtering in Maintaining Safe Alignment of LLMs

ACL ARR 2025 February Submission2459 Authors

Abstract: While Large Language Models (LLMs) have shown significant advancements in performance, various jailbreak attacks have posed growing safety and ethical risks. Malicious users often exploit adversarial context to deceive LLMs, prompting them to generate responses to harmful queries. In this study, we propose a new defense mechanism called Context Filtering model—an input pre-processing method designed to filter out untrustworthy and unreliable context while identifying the primary prompts containing the real user intent to uncover concealed malicious intent. Given that enhancing the safety of LLMs often compromises their helpfulness, potentially affecting the experience of benign users, our method aims to improve the safety of the LLMs while preserving their original performance. We evaluate the effectiveness of our model in defending against jailbreak attacks through comparative analysis, comparing our approach with state-of-the-art defense mechanisms against three different attacks and assessing the helpfulness of LLMs under these defenses. Our model demonstrates its ability to reduce the Attack Success Rates of jailbreak attacks by up to 84% while maintaining the original LLMs' performance, achieving state-of-the-art Safety and Helpfulness Product results. Notably, our model is a plug-and-play method that can be applied to all LLMs, including both white-box and black-box models, to enhance their safety without requiring any fine-tuning of the models themselves. We will make our model publicly available for research purposes.

Paper Type: Long

Research Area: Language Modeling

Research Area Keywords: security and privacy, prompting, fine-tuning

Contribution Types: Model analysis & interpretability

Languages Studied: English

Submission Number: 2459