ExpGuard: LLM Content Moderation in Specialized Domains

TL;DR

ExpGuard is a specialized safety guardrail for large language models, focusing on domain-specific content moderation in finance, medical, and legal sectors, with a new dataset and robust evaluation methods.

Contribution

We introduce ExpGuard, a domain-specific guardrail model and dataset, enhancing safety and robustness of LLMs against technical and adversarial content in specialized fields.

Findings

ExpGuard outperforms existing models in domain-specific prompt and response classification.

The dataset ExpGuardMix enables effective training and evaluation of specialized guardrails.

ExpGuard demonstrates high resilience to adversarial attacks in targeted domains.

Abstract

With the growing deployment of large language models (LLMs) in real-world applications, establishing robust safety guardrails to moderate their inputs and outputs has become essential to ensure adherence to safety policies. Current guardrail models predominantly address general human-LLM interactions, rendering LLMs vulnerable to harmful and adversarial content within domain-specific contexts, particularly those rich in technical jargon and specialized concepts. To address this limitation, we introduce ExpGuard, a robust and specialized guardrail model designed to protect against harmful prompts and responses across financial, medical, and legal domains. In addition, we present ExpGuardMix, a meticulously curated dataset comprising 58,928 labeled prompts paired with corresponding refusal and compliant responses, from these specific sectors. This dataset is divided into two subsets: ExpGuardTrain, for model training, and ExpGuardTest, a high-quality test set annotated by domain experts to evaluate model robustness against technical and domain-specific content. Comprehensive evaluations conducted on ExpGuardTest and eight established public benchmarks reveal that ExpGuard delivers competitive performance across the board while demonstrating exceptional resilience to domain-specific adversarial attacks, surpassing state-of-the-art models such as WildGuard by up to 8.9% in prompt classification and 15.3% in response classification. To encourage further research and development, we open-source our code, data, and model, enabling adaptation to additional domains and supporting the creation of increasingly robust guardrail models.