Article citationsMore>>

Tunstall, L., Beeching, E., Rasul, K., Belkada, Y., Huang, S., von Werra, L., Fourrier, C., Habib, N., Sarrazin, N., Sanseviero, O., Rush, A.M. and Wolf, T. (2023) Zephyr: Direct Distillation of LM Alignment. arXiv preprint arXiv:2310.16944.

has been cited by the following article:

• TITLE: GUARDIAN: A Multi-Tiered Defense Architecture for Thwarting Prompt Injection Attacks on LLMs

  AUTHORS: Parijat Rai, Saumil Sood, Vijay K. Madisetti, Arshdeep Bahga

  KEYWORDS: Large Language Models (LLMs), Adversarial Attack, Prompt Injection, Filter Defense, Artificial Intelligence, Machine Learning, Cybersecurity

  JOURNAL NAME: Journal of Software Engineering and Applications, Vol.17 No.1, January 23, 2024

  ABSTRACT: This paper introduces a novel multi-tiered defense architecture to protect language models from adversarial prompt attacks. We construct adversarial prompts using strategies like role emulation and manipulative assistance to simulate real threats. We introduce a comprehensive, multi-tiered defense framework named GUARDIAN (Guardrails for Upholding Ethics in Language Models) comprising a system prompt filter, pre-processing filter leveraging a toxic classifier and ethical prompt generator, and pre-display filter using the model itself for output screening. Extensive testing on Meta’s Llama-2 model demonstrates the capability to block 100% of attack prompts. The approach also auto-suggests safer prompt alternatives, thereby bolstering language model security. Quantitatively evaluated defense layers and an ethical substitution mechanism represent key innovations to counter sophisticated attacks. The integrated methodology not only fortifies smaller LLMs against emerging cyber threats but also guides the broader application of LLMs in a secure and ethical manner.