Efficient Jailbreak Mitigation Using Semantic Linear Classification in a Multi-Staged Pipeline
This addresses security vulnerabilities in LLM-based systems with an efficient, incremental improvement over existing defenses.
The paper tackles prompt injection and jailbreak attacks on large language models by proposing a multi-stage defense pipeline with a semantic filter using TF-IDF and Linear SVM, achieving 93.4% accuracy and reducing latency from 450s to 47s compared to prior methods.
Prompt injection and jailbreaking attacks pose persistent security challenges to large language model (LLM)-based systems. We present an efficient and systematically evaluated defense architecture that mitigates these threats through a lightweight, multi-stage pipeline. Its core component is a semantic filter based on text normalization, TF-IDF representations, and a Linear SVM classifier. Despite its simplicity, this module achieves 93.4% accuracy and 96.5% specificity on held-out data, substantially reducing attack throughput while incurring negligible computational overhead. Building on this efficient foundation, the full pipeline integrates complementary detection and mitigation mechanisms that operate at successive stages, providing strong robustness with minimal latency. In comparative experiments, our SVM-based configuration improves overall accuracy from 35.1% to 93.4% while reducing average time to completion from approximately 450s to 47s, yielding over 10 times lower latency than ShieldGemma. These results demonstrate that the proposed design simultaneously advances defensive precision and efficiency, addressing a core limitation of current model-based moderators. Evaluation across a curated corpus of over 30,000 labeled prompts, including benign, jailbreak, and application-layer injections, confirms that staged, resource-efficient defenses can robustly secure modern LLM-driven applications.