arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2402.01822cs.CLcs.AI
收藏

构建大型语言模型的护栏

Building Guardrails for Large Language Models

University of Liverpool(利物浦大学);Key Laboratory of System Software (Chinese Academy of Sciences)(中国科学院系统软件重点实验室);State Key Laboratory of Computer Science(计算机科学国家重点实验室);Institute of Software, Chinese Academy of Sciences(中国科学院软件研究所);University of Warwick(华威大学);Institute of Digital Technologies, Loughborough University London(拉夫堡大学伦敦数字技术研究所)

Yi Dong, Ronghui Mu, Gaojie Jin, Yi Qi, Jinwei Hu, Xingyu Zhao, Jie Meng, Wenjie Ruan, Xiaowei Huang

AI总结:

本文提出系统性构建LLM护栏的方法,通过多学科协作明确需求、神经符号实现及验证测试,以应对LLM应用中的风险。

摘要:

随着大型语言模型(LLMs)日益融入我们的日常生活,识别并缓解其风险变得至关重要,尤其是当这些风险可能对人类用户和社会产生深远影响时。护栏(Guardrails)作为一种过滤LLMs输入或输出的技术,已成为核心的安全保障技术。本立场文件深入审视了当前的开源解决方案(Llama Guard、Nvidia NeMo、Guardrails AI),并讨论了构建更完整解决方案所面临的挑战与路径。基于先前研究的坚实证据,我们倡导一种系统性的方法来构建LLMs护栏,该方法基于对各种LLMs应用场景中不同上下文的全面考虑。我们提议通过多学科团队的协作采用社会技术方法,以明确精确的技术需求;探索先进的神经符号实现,以应对需求的复杂性;并开发验证与测试机制,以确保最终产品的最高质量。

英文摘要:

As Large Language Models (LLMs) become more integrated into our daily lives, it is crucial to identify and mitigate their risks, especially when the risks can have profound impacts on human users and societies. Guardrails, which filter the inputs or outputs of LLMs, have emerged as a core safeguarding technology. This position paper takes a deep look at current open-source solutions (Llama Guard, Nvidia NeMo, Guardrails AI), and discusses the challenges and the road towards building more complete solutions. Drawing on robust evidence from previous research, we advocate for a systematic approach to construct guardrails for LLMs, based on comprehensive consideration of diverse contexts across various LLMs applications. We propose employing socio-technical methods through collaboration with a multi-disciplinary team to pinpoint precise technical requirements, exploring advanced neural-symbolic implementations to embrace the complexity of the requirements, and developing verification and testing to ensure the utmost quality of the final product.

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。

↑