Towards Assurance of LLM Adversarial Robustness using Ontology-Driven Argumentation

2024/10/10 by Tomas Bueno Momčilović, Momcilovic, Tomas Bueno, Beat Buesser +7
Computer Science · #Adversarial Robustness in Machine Learning #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences

paper · pdf · doi:10.48550/arxiv.2410.07962

openalex publication_date 2024/10/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Despite the impressive adaptability of large language models (LLMs), challenges remain in ensuring their security, transparency, and interpretability. Given their susceptibility to adversarial attacks, LLMs need to be defended with an evolving combination of adversarial training and guardrails. However, managing the implicit and heterogeneous knowledge for continuously assuring robustness is difficult. We introduce a novel approach for assurance of the adversarial robustness of LLMs based on formal argumentation. Using ontologies for formalization, we structure state-of-the-art attacks and defenses, facilitating the creation of a human-readable assurance case, and a machine-readable representation. We demonstrate its application with examples in English language and code translation tasks, and provide implications for theory and practice, by targeting engineers, data scientists, users, and auditors.

Related

• Developing Assurance Cases for Adversarial Robustness and Regulatory Compliance in LLMs
• Towards Assuring EU AI Act Compliance and Adversarial Robustness of LLMs
• Advancing NLP Security by Leveraging LLMs as Adversarial Engines
• A Survey of Attacks on Large Language Models
• LLMs Cannot Reliably Judge (Yet?): A Comprehensive Assessment on the Robustness of LLM-as-a-Judge