LoRA-Guard: Parameter-Efficient Guardrail Adaptation for Content Moderation of Large Language Models

TL;DR

LoRA-Guard is a parameter-efficient method enabling on-device content moderation for large language models by sharing knowledge between models with minimal resource overhead.

Contribution

It introduces a novel low-rank adapter-based approach for guardrail adaptation that significantly reduces parameter overhead while maintaining accuracy.

Findings

Outperforms existing methods with 100-1000x fewer parameters

Maintains accuracy on content moderation tasks

Enables on-device deployment of LLM guardrails

Abstract

Guardrails have emerged as an alternative to safety alignment for content moderation of large language models (LLMs). Existing model-based guardrails have not been designed for resource-constrained computational portable devices, such as mobile phones, more and more of which are running LLM-based applications locally. We introduce LoRA-Guard, a parameter-efficient guardrail adaptation method that relies on knowledge sharing between LLMs and guardrail models. LoRA-Guard extracts language features from the LLMs and adapts them for the content moderation task using low-rank adapters, while a dual-path design prevents any performance degradation on the generative task. We show that LoRA-Guard outperforms existing approaches with 100-1000x lower parameter overhead while maintaining accuracy, enabling on-device content moderation.