arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.08946cs.LGcs.CLcs.CY
收藏

可用的可解释人工智能(Usable XAI):大语言模型时代利用可解释性的10项策略

Usable XAI: 10 Strategies Towards Exploiting Explainability in the LLM Era

University of Georgia(佐治亚大学);New Jersey Institute of Technology(新泽西理工学院);University of Virginia(弗吉尼亚大学);Wake Forest University(威克森林大学);King Abdullah University of Science and Technology(国王阿卜杜勒阿齐兹大学);Amazon(亚马逊)

Xuansheng Wu, Haiyan Zhao, Yaochen Zhu, Yucheng Shi, Fan Yang, Lijie Hu, Tianming Liu, Xiaoming Zhai, Wenlin Yao, Jundong Li, Mengnan Du, Ninghao Liu

AI总结:

该研究针对大语言模型时代XAI的应用转型需求,提出“可用XAI”概念,梳理10项双向赋能的可解释性策略及技术挑战,辅以案例与开源代码,推动XAI在LLM场景的实用化。

摘要:

可解释人工智能(XAI)指的是为人类提供可理解的AI模型运作洞察的技术。近年来,XAI的研究重点正扩展到大语言模型(LLMs)的可解释性上。这一延伸要求XAI方法论发生重大转变,原因有二:其一,现有许多XAI方法因LLMs的复杂性和高阶能力无法直接应用于这类模型;其二,随着LLMs被越来越多地部署到各类应用中,XAI的角色从仅仅打开“黑箱”,转变为在真实场景中主动提升LLMs的生产力与适用性。与此同时,LLMs的对话和生成能力也能反过来增强XAI。因此,本文在LLMs语境下提出“可用XAI(Usable XAI)”概念,分析了两个方向:(1)XAI如何解释并改进基于LLM的AI系统;(2)如何借助LLMs优化XAI技术。本文提出10项策略,介绍每项策略的核心技术并讨论相关挑战,还提供了案例研究展示如何获取和利用解释结果。本文所用代码可访问:https://github.com/JacksonWuxs/UsableXAI_LLM。

英文摘要:

Explainable AI (XAI) refers to techniques that provide human-understandable insights into the workings of AI models. Recently, the focus of XAI is being extended toward explaining Large Language Models (LLMs). This extension calls for a significant transformation in the XAI methodologies for two reasons. First, many existing XAI methods cannot be directly applied to LLMs due to their complexity and advanced capabilities. Second, as LLMs are increasingly deployed in diverse applications, the role of XAI shifts from merely opening the ``black box'' to actively enhancing the productivity and applicability of LLMs in real-world settings. Meanwhile, the conversation and generation abilities of LLMs can reciprocally enhance XAI. Therefore, in this paper, we introduce Usable XAI in the context of LLMs by analyzing (1) how XAI can explain and improve LLM-based AI systems and (2) how XAI techniques can be improved by using LLMs. We introduce 10 strategies, introducing the key techniques for each and discussing their associated challenges. We also provide case studies to demonstrate how to obtain and leverage explanations. The code used in this paper can be found at: https://github.com/JacksonWuxs/UsableXAI_LLM.

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。

↑