发表机构
The University of Tokyo; CNRS（东京大学; 法国国家科学研究中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究在范畴论框架下，通过分析COGS数据集的21种泛化类型，以结构识别替代模型训练来诊断组合泛化的可容许性，明确训练语料库的许可范围。
AI 中文摘要
组合泛化通常通过模型准确率评估，我们转而探究哪些结构或词汇识别能使保留的COGS示例符合训练中观测到的结构。句子被表示为从句法地址到词汇标记的函子，选择性坍缩会诱导Kan扩张以传播观测到的关联。在21种COGS泛化类型中，可容许性遵循不同的识别轮廓，而残余失败则对应不被支持的结构模板。这些数据层面的诊断刻画了训练语料库在指定识别下许可的内容，无需训练预测模型。
英文摘要
Compositional generalization is usually evaluated through model accuracy. We instead ask which structural or lexical identifications make held-out COGS examples admissible from the structures observed in training. Sentences are represented as functors from syntactic addresses to lexical tokens, and selective collapses induce Kan extensions that propagate observed associations. Across 21 COGS generalization types, admissibility follows distinct identification profiles, while residual failures separate unsupported structural templates. These data-side diagnoses characterize what the training corpus licenses under specified identifications, without training a predictive model.
Comments11 pages
