大型语言模型中的中文知识校正基准测试

Tianhe Lu, Jizhan Fang, Yunzhi Yao, Xin Xu, Ningyu Zhang, Huajun Chen
{"title":"大型语言模型中的中文知识校正基准测试","authors":"Tianhe Lu, Jizhan Fang, Yunzhi Yao, Xin Xu, Ningyu Zhang, Huajun Chen","doi":"arxiv-2409.05806","DOIUrl":null,"url":null,"abstract":"While Large Language Models (LLMs) exhibit remarkable generative\ncapabilities, they are not without flaws, particularly in the form of\nhallucinations. This issue is even more pronounced when LLMs are applied to\nspecific languages and domains. For example, LLMs may generate nonsense\ninformation when handling Chinese ancient poetry, proverbs, or idioms, owing to\nthe lack of specific knowledge. To this end, this paper introduces a benchmark\nfor rectifying Chinese knowledge in LLMs via knowledge editing. Specifically,\nwe introduce a new Chinese dataset, CKnowEdit, by collecting seven type of\nknowledge from various sources, including classical texts, idioms, and content\nfrom Baidu Tieba Ruozhiba, thereby accounting for the unique polyphony,\nantithesis, and logical constructs inherent in the Chinese language. Through\nthe analysis of this dataset, we uncover the challenges faced by current LLMs\nin mastering Chinese. Furthermore, our evaluation of state-of-the-art knowledge\nediting techniques on this dataset unveil the substantial scope for advancement\nin the rectification of Chinese knowledge. Code and dataset are available at\nhttps://github.com/zjunlp/EasyEdit.","PeriodicalId":501281,"journal":{"name":"arXiv - CS - Information Retrieval","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Benchmarking Chinese Knowledge Rectification in Large Language Models\",\"authors\":\"Tianhe Lu, Jizhan Fang, Yunzhi Yao, Xin Xu, Ningyu Zhang, Huajun Chen\",\"doi\":\"arxiv-2409.05806\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"While Large Language Models (LLMs) exhibit remarkable generative\\ncapabilities, they are not without flaws, particularly in the form of\\nhallucinations. This issue is even more pronounced when LLMs are applied to\\nspecific languages and domains. For example, LLMs may generate nonsense\\ninformation when handling Chinese ancient poetry, proverbs, or idioms, owing to\\nthe lack of specific knowledge. To this end, this paper introduces a benchmark\\nfor rectifying Chinese knowledge in LLMs via knowledge editing. Specifically,\\nwe introduce a new Chinese dataset, CKnowEdit, by collecting seven type of\\nknowledge from various sources, including classical texts, idioms, and content\\nfrom Baidu Tieba Ruozhiba, thereby accounting for the unique polyphony,\\nantithesis, and logical constructs inherent in the Chinese language. Through\\nthe analysis of this dataset, we uncover the challenges faced by current LLMs\\nin mastering Chinese. Furthermore, our evaluation of state-of-the-art knowledge\\nediting techniques on this dataset unveil the substantial scope for advancement\\nin the rectification of Chinese knowledge. Code and dataset are available at\\nhttps://github.com/zjunlp/EasyEdit.\",\"PeriodicalId\":501281,\"journal\":{\"name\":\"arXiv - CS - Information Retrieval\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Information Retrieval\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.05806\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Information Retrieval","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.05806","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

虽然大语言模型(LLMs)具有非凡的生成能力,但也并非没有缺陷,尤其是在幻觉方面。当 LLM 应用于特定语言和领域时,这一问题就更加突出。例如,在处理中国古诗词、谚语或成语时,由于缺乏特定的知识,LLMs 可能会产生无意义的信息。为此,本文介绍了一种通过知识编辑纠正 LLM 中中文知识的基准。具体来说,我们引入了一个新的中文数据集--CKnowEdit,从各种来源收集七类知识,包括古典文本、成语和百度铁算盘开奖结果中的内容,从而考虑到中文特有的复音、对立和逻辑构造。通过对该数据集的分析,我们揭示了当前法律硕士在掌握中文时所面临的挑战。此外,我们在该数据集上对最先进的知识编辑技术进行了评估,从而揭示了中文知识修正的巨大进步空间。代码和数据集可在https://github.com/zjunlp/EasyEdit。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Benchmarking Chinese Knowledge Rectification in Large Language Models
While Large Language Models (LLMs) exhibit remarkable generative capabilities, they are not without flaws, particularly in the form of hallucinations. This issue is even more pronounced when LLMs are applied to specific languages and domains. For example, LLMs may generate nonsense information when handling Chinese ancient poetry, proverbs, or idioms, owing to the lack of specific knowledge. To this end, this paper introduces a benchmark for rectifying Chinese knowledge in LLMs via knowledge editing. Specifically, we introduce a new Chinese dataset, CKnowEdit, by collecting seven type of knowledge from various sources, including classical texts, idioms, and content from Baidu Tieba Ruozhiba, thereby accounting for the unique polyphony, antithesis, and logical constructs inherent in the Chinese language. Through the analysis of this dataset, we uncover the challenges faced by current LLMs in mastering Chinese. Furthermore, our evaluation of state-of-the-art knowledge editing techniques on this dataset unveil the substantial scope for advancement in the rectification of Chinese knowledge. Code and dataset are available at https://github.com/zjunlp/EasyEdit.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
Decoding Style: Efficient Fine-Tuning of LLMs for Image-Guided Outfit Recommendation with Preference Retrieve, Annotate, Evaluate, Repeat: Leveraging Multimodal LLMs for Large-Scale Product Retrieval Evaluation Active Reconfigurable Intelligent Surface Empowered Synthetic Aperture Radar Imaging FLARE: Fusing Language Models and Collaborative Architectures for Recommender Enhancement Basket-Enhanced Heterogenous Hypergraph for Price-Sensitive Next Basket Recommendation
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1