AfriGuard: Safety and Robust Refusal For African Languages

📄 arXiv: 2610.26474v1 📥 PDF

作者: Israel Abebe Azime, Jesujoba O. Alabi, Iffat Maab, Alexander D. Cai, Tadesse Destaw Belay, Atnafu Lambebo Tonja, Jonas Kgomo, Folasade Alabi, Salomey Osei, Saminu Muhammad Aliyu, Nkechinyere Faith Aguobi, Bontu Fufa Balcha, Blessing Sibanda, Davis David, Mouhamadane Mboup, Daud Abolade, Neo Putini, Louis Powell, Kiant'e Brantley, Philipp Slusallek, Dietrich Klakow, Julia Kreutzer

分类: cs.CL

发布日期: 2026-10-19


💡 一句话要点

提出AfriGuard以解决非洲语言安全响应问题

🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 大型语言模型 安全响应 多语言数据集 有害意图识别 拒绝检测器 文化背景 非洲语言 监督微调

📋 核心要点

  1. 现有的语言模型在多语言和文化背景下对安全风险的识别和响应存在不足,尤其是在非洲语言中。
  2. AfriGuard是一个多语言数据集,旨在研究不同文化背景下的有害意图和拒绝行为,提升模型的安全响应能力。
  3. 实验表明,使用AfriGuard的模型在安全性能上显著提升,拒绝检测器的性能提高了43.9%。

📝 摘要(中文)

大型语言模型(LLMs)在与用户互动时,通常被训练以适当地响应有害或有毒内容。然而,在多样的语言和文化背景下,确保适当的响应变得更加复杂。本文介绍了AfriGuard,这是一个以文化为基础的多语言数据集,用于研究10种非洲语言和英语中的有害意图、安全类别理解和拒绝行为。研究发现,现有的守护模型在安全意图和安全类别分类上表现不佳,且对不安全提示的响应往往不够恰当。通过AfriGuard,本文开发了高效的守护模型,其性能与最大30倍规模的模型相当,并且引入的拒绝检测器在现有检测器基础上提升了43.9%的平均性能。

🔬 方法详解

问题定义:本文旨在解决大型语言模型在多语言和文化背景下对安全风险的识别和响应不足的问题。现有方法在处理非洲语言时,往往无法准确理解和响应有害内容。

核心思路:论文提出了AfriGuard数据集,通过对多种非洲语言的有害意图和拒绝行为进行深入研究,旨在提升模型的安全性和适应性。设计思路强调文化背景对安全响应的影响。

技术框架:整体架构包括数据集构建、模型训练和评估三个主要阶段。数据集涵盖多种语言的安全意图和拒绝行为,模型则通过监督微调来提升性能。

关键创新:最重要的创新在于引入了AfriGuard数据集和高效的拒绝检测器,这些方法在处理多语言安全响应时表现出色,显著提升了模型的安全性能。

关键设计:在模型训练中,采用了特定的损失函数和网络结构,以优化对安全意图的分类和拒绝行为的识别,确保模型在多语言环境下的有效性。具体参数设置和网络结构细节在论文中进行了详细描述。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示,使用AfriGuard的模型在安全意图和安全类别分类上表现优于现有的开源和闭源LLMs,且拒绝检测器的平均性能提升了43.9%。这些结果表明,AfriGuard在提升多语言安全响应方面具有显著效果。

🎯 应用场景

该研究的潜在应用领域包括社交媒体监控、在线客服系统和内容审核等。通过提升对多语言环境中有害内容的识别和响应能力,AfriGuard能够帮助构建更安全的AI系统,促进不同文化背景下的用户安全体验。

📄 摘要(原文)

Large language models (LLMs) designed to interact with users are typically trained to respond appropriately to harmful or toxic content. However, ensuring appropriate responses becomes more challenging in diverse linguistic and cultural contexts, where safety risks may be expressed or interpreted differently. Building safe AI systems therefore requires detecting safety risks, understanding their nature, and responding appropriately, while accounting for linguistic and cultural variations that add complexity to each stage. In this work, we introduce AfriGuard, a culturally grounded multilingual dataset for studying harmful intent, safety-category understanding, and refusal behavior across 10 African languages and English. Our findings reveal that guardrail models underperform both open- and closed-source LLMs on safety-intent and safety-category classification, while models show limited understanding of appropriate safety behavior, scoring below 60% F1. We further find that models often fail to respond appropriately to unsafe prompts not because they explicitly refuse, but because they produce linguistically inadequate responses that result in incidental non-assistance rather than proper refusals. Finally, using AfriGuard, we develop efficient guardrail models that achieve performance competitive with models up to 30x larger, introduce refusal detectors that improve average performance by 43.9% over existing detectors, and demonstrate that supervised fine-tuning can substantially improve overall safety performance.