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 2104.08704cs.CLcs.AI
收藏

用于自由文本生成的Token级无参考幻觉检测基准

A Token-level Reference-free Hallucination Detection Benchmark for Free-form Text Generation

Peking University(北京大学);Tencent Cloud Xiaowei(腾讯云小微);Meta AI;Microsoft Corporation(微软公司)

Tianyu Liu, Yizhe Zhang, Chris Brockett, Yi Mao, Zhifang Sui, Weizhu Chen, Bill Dolan

摘要:

像GPT-3这样的大型预训练生成模型经常会出现不存在或不正确的内容,这削弱了它们在实际应用中的潜在优势。现有的工作通常试图检测这些幻觉的基础上相应的甲骨文参考在一个句子或文档的水平。然而,对于许多自由形式的文本生成应用,地面实况参考可能不容易获得,并且句子级或文档级检测可能无法提供将实时的防止错误内容的细粒度信号。作为解决这些问题的第一步,我们提出了一种新的令牌级,无参考的幻觉检测任务和一个相关的注释数据集命名为HaDes(HAllucination检测数据集)。为了创建这个数据集,我们首先扰动了从英语维基百科中提取的大量文本片段,然后用众包注释验证这些文本片段。为了减轻标注过程中的标签不平衡,我们采用了迭代模型在环策略。我们进行全面的数据分析并创建多个基线模型。

英文摘要:

Large pretrained generative models like GPT-3 often suffer from hallucinating non-existent or incorrect content, which undermines their potential merits in real applications. Existing work usually attempts to detect these hallucinations based on a corresponding oracle reference at a sentence or document level. However ground-truth references may not be readily available for many free-form text generation applications, and sentence- or document-level detection may fail to provide the fine-grained signals that would prevent fallacious content in real time. As a first step to addressing these issues, we propose a novel token-level, reference-free hallucination detection task and an associated annotated dataset named HaDes (HAllucination DEtection dataSet). To create this dataset, we first perturb a large number of text segments extracted from English language Wikipedia, and then verify these with crowd-sourced annotations. To mitigate label imbalance during annotation, we utilize an iterative model-in-loop strategy. We conduct comprehensive data analyses and create multiple baseline models.

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。

↑