发表机构
University of Cincinnati（辛辛那提大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对现有法律含义保留指标检测的缺陷，提出LexFlip解离诊断工具，测试发现多数语义指标对法律效力反转的微小扰动不敏感，单纯长度特征在法律含义评估任务中表现优于多数语义指标。
AI 中文摘要
简化后的法律条款是否仍能表达与原条款相同的含义？当前使用的检测方法无法确认这一点：要求相同的条款对得分最高、不相关的条款对得分最低，会将词汇重叠度与法律效力绑定，因此任何基于词元重叠的单调函数都能满足这两个要求。我们的解决方案是一种解离设计，即固定表面形式但改变法律效力。我们发布了LexFlip，这是373个对魁北克法语成文法的微小扰动，可在保留93%词元的同时反转法律效力，配套工具可对指标、回归模型和提示式法官进行评分。我们测试的7种嵌入指标和BERTScore指标，在这类编辑上的得分仅占“相同-不相关”区间的0.022至0.039，而双向NLI家族的这一比例为0.670，是唯一会被“相同对检测”排除的家族。在FrJudge数据集上，相对于人类测得的上限相关系数r=0.597，单纯的长度特征胜过所有语义指标，且具有我们测得的最低误差幅度。
英文摘要
Does a simplified legal clause still say what the original said? The checks in current use cannot establish that it does: requiring an identical pair to score highest and an unrelated pair lowest moves lexical overlap and legal force together, so any monotone function of token overlap satisfies both. Our remedy is a dissociation, an item holding surface form fixed while legal force moves. We release LexFlip, 373 minimal perturbations of Quebec statutory French that reverse legal force while preserving 0.93 of the tokens, with a harness scoring metrics, regressors and prompted judges alike. The seven embedding and BERTScore metrics we test spend only 0.022 to 0.039 of their identical-to-unrelated range on such an edit, against 0.670 for bidirectional NLI, the one family the identical-pair check would disqualify. On FrJudge, against a measured human ceiling of r=0.597, a bare length feature outscores every semantic metric and has the lowest margin we measure.
Comments6 pages, 2 figures, 3 tables
