Towards Assurance of LLM Adversarial Robustness using Ontology-Driven Argumentation
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866912067625156608 |
|---|---|
| author | Momcilovic, Tomas Bueno Buesser, Beat Zizzo, Giulio Purcell, Mark Balta, Dian |
| author_facet | Momcilovic, Tomas Bueno Buesser, Beat Zizzo, Giulio Purcell, Mark Balta, Dian |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_07962 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Towards Assurance of LLM Adversarial Robustness using Ontology-Driven Argumentation Momcilovic, Tomas Bueno Buesser, Beat Zizzo, Giulio Purcell, Mark Balta, Dian Artificial Intelligence 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. |
| title | Towards Assurance of LLM Adversarial Robustness using Ontology-Driven Argumentation |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2410.07962 |