arXiv 2404.066642024-04-10cs.CLcs.AIcs.HC

CulturalTeaming：用于挑战大语言模型（缺失的）多元文化知识的AI辅助交互式红队测试

CulturalTeaming: AI-Assisted Interactive Red-Teaming for Challenging LLMs' (Lack of) Multicultural Knowledge

University of Washington（华盛顿大学）；Allen Institute for Artificial Intelligence（艾伦人工智能研究所）；Carnegie Mellon University（卡内基梅隆大学）；University of British Columbia（不列颠哥伦比亚大学）；Vector Institute for AI（矢量人工智能研究所）；Paul G. Allen School of Computer Science & Engineering, University of Washington（华盛顿大学保罗·G·艾伦计算机科学与工程学院）

Yu Ying Chiu, Liwei Jiang, Maria Antoniak, Chan Young Park, Shuyue Stella Li, Mehar Bhatia, Sahithya Ravi, Yulia Tsvetkov, Vered Shwartz, Yejin Choi

备注：: Preprint (under review)

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对现有大语言模型多元文化知识评估方法的不足，本研究提出人机协作的交互式红队系统CulturalTeaming，构建了高质量评估数据集CULTURALBENCH-V0.1，发现现代大语言模型多元文化能力存在明显差距，且更高程度AI辅助可提升问题难度与用户创造力感知。

摘要：

前沿大语言模型（LLM）由文化背景存在偏差的研究人员和从业者开发，且训练于来源存在偏差的数据集之上。然而，当前的基准构建方法无法有效评估大语言模型的多元文化知识（缺失）情况。现有的多元文化评估主要依赖成本高昂且受限的人工标注，或是可能已经过时的互联网资源，因此难以捕捉文化规范的复杂性、动态性和多样性。由大语言模型生成的基准虽然前景可观，但存在传播其本应测量的偏见的风险。为了协同人类标注者的创造力与专业文化知识，以及基于大语言模型的自动化的可扩展性和可标准化性，我们提出了CulturalTeaming——这是一个交互式红队测试系统，利用人机协作构建真正具有挑战性的评估数据集，用于评估大语言模型的多元文化知识，同时提升标注者的能力和体验。我们的研究表明，CulturalTeaming的多种AI辅助模式能够支持标注者以游戏化的方式创建现代大语言模型无法答对的文化问题。重要的是，更高程度的AI辅助（例如大语言模型生成的修改提示）能够让用户创建出难度更高的问题，同时提升用户对自身创造力的感知，这为在现代评估数据集构建流程中引入更强的AI辅助带来了启示。通过一系列时长1小时的工作坊环节，我们收集了CULTURALBENCH-V0.1——这是一个包含用户红队测试尝试的紧凑但高质量的评估数据集，不同系列的现代大语言模型在该数据集上的准确率在37.7%到72.2%之间，揭示出大语言模型在多元文化能力方面存在显著差距。

英文摘要：

Frontier large language models (LLMs) are developed by researchers and practitioners with skewed cultural backgrounds and on datasets with skewed sources. However, LLMs' (lack of) multicultural knowledge cannot be effectively assessed with current methods for developing benchmarks. Existing multicultural evaluations primarily rely on expensive and restricted human annotations or potentially outdated internet resources. Thus, they struggle to capture the intricacy, dynamics, and diversity of cultural norms. LLM-generated benchmarks are promising, yet risk propagating the same biases they are meant to measure. To synergize the creativity and expert cultural knowledge of human annotators and the scalability and standardizability of LLM-based automation, we introduce CulturalTeaming, an interactive red-teaming system that leverages human-AI collaboration to build truly challenging evaluation dataset for assessing the multicultural knowledge of LLMs, while improving annotators' capabilities and experiences. Our study reveals that CulturalTeaming's various modes of AI assistance support annotators in creating cultural questions, that modern LLMs fail at, in a gamified manner. Importantly, the increased level of AI assistance (e.g., LLM-generated revision hints) empowers users to create more difficult questions with enhanced perceived creativity of themselves, shedding light on the promises of involving heavier AI assistance in modern evaluation dataset creation procedures. Through a series of 1-hour workshop sessions, we gather CULTURALBENCH-V0.1, a compact yet high-quality evaluation dataset with users' red-teaming attempts, that different families of modern LLMs perform with accuracy ranging from 37.7% to 72.2%, revealing a notable gap in LLMs' multicultural proficiency.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。