A Domain-Based Taxonomy of Jailbreak Vulnerabilities in Large Language Models
This work addresses jailbreak vulnerabilities in LLMs, offering a domain-based perspective that is more insightful than conventional classifications, though it is incremental as it builds on existing understanding without introducing new mitigation methods.
The paper tackles the problem of jailbreak vulnerabilities in large language models by introducing a novel taxonomy based on training domains, categorizing attacks into mismatched generalization, competing objectives, adversarial robustness, and mixed attacks to provide deeper insights into model deficiencies.
The study of large language models (LLMs) is a key area in open-world machine learning. Although LLMs demonstrate remarkable natural language processing capabilities, they also face several challenges, including consistency issues, hallucinations, and jailbreak vulnerabilities. Jailbreaking refers to the crafting of prompts that bypass alignment safeguards, leading to unsafe outputs that compromise the integrity of LLMs. This work specifically focuses on the challenge of jailbreak vulnerabilities and introduces a novel taxonomy of jailbreak attacks grounded in the training domains of LLMs. It characterizes alignment failures through generalization, objectives, and robustness gaps. Our primary contribution is a perspective on jailbreak, framed through the different linguistic domains that emerge during LLM training and alignment. This viewpoint highlights the limitations of existing approaches and enables us to classify jailbreak attacks on the basis of the underlying model deficiencies they exploit. Unlike conventional classifications that categorize attacks based on prompt construction methods (e.g., prompt templating), our approach provides a deeper understanding of LLM behavior. We introduce a taxonomy with four categories -- mismatched generalization, competing objectives, adversarial robustness, and mixed attacks -- offering insights into the fundamental nature of jailbreak vulnerabilities. Finally, we present key lessons derived from this taxonomic study.