Efficient Detection of Toxic Prompts in Large Language Models

Large language models (LLMs) like ChatGPT and Gemini have significantly advanced natural language processing, enabling various applications such as chatbots and automated content generation. However, these models can be exploited by malicious individuals who craft toxic prompts to elicit harmful or...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE/ACM International Conference on Automated Software Engineering : [proceedings] S. 455 - 467
Hauptverfasser: Liu, Yi, Yu, Junzhe, Sun, Huijia, Shi, Ling, Deng, Gelei, Chen, Yuqi, Liu, Yang
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: ACM 27.10.2024
Schlagworte:
Accuracy
ai security
Chatbots
detection
Feature extraction
Glass box
jailbreak
large language model
Large language models
llm
llm security
Real-time systems
Safety
Scalability
Software engineering
toxic prompt
Vectors
ISSN:2643-1572
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Large language models (LLMs) like ChatGPT and Gemini have significantly advanced natural language processing, enabling various applications such as chatbots and automated content generation. However, these models can be exploited by malicious individuals who craft toxic prompts to elicit harmful or unethical responses. These individuals often employ jailbreaking techniques to bypass safety mechanisms, highlighting the need for robust toxic prompt detection methods. Existing detection techniques, both blackbox and whitebox, face challenges related to the diversity of toxic prompts, scalability, and computational efficiency. In response, we propose ToxicDetector, a lightweight greybox method designed to efficiently detect toxic prompts in LLMs. ToxicDetector leverages LLMs to create toxic concept prompts, uses embedding vectors to form feature vectors, and employs a Multi-Layer Perceptron (MLP) classifier for prompt classification. Our evaluation on various versions of the LLama models, Gemma-2, and multiple datasets demonstrates that ToxicDetector achieves a high accuracy of 96.39% and a low false positive rate of 2.00%, outperforming state-of-the-art methods. Additionally, ToxicDetector's processing time of 0.0780 seconds per prompt makes it highly suitable for real-time applications. ToxicDetector achieves high accuracy, efficiency, and scalability, making it a practical method for toxic prompt detection in LLMs.
ISSN:2643-1572
DOI:10.1145/3691620.3695018