直接和间接证据共同影响语言模型中的与格交替偏好
Both Direct and Indirect Evidence Contribute to Dative Alternation Preferences in Language Models
AI总结：
本研究通过受控养育范式和系统化操纵输入，探索英语与格交替中长度和生命性对语言模型偏好的影响，发现这些偏好源于直接和间接证据的共同作用。
摘要：
语言模型（LMs）在许多句法现象上表现出类似人类的偏好，但这些偏好可归因于对这些现象的直接接触还是语言的更普遍属性，目前尚不清楚。我们使用英语与格交替（DO：“gave Y the X” 对比 PO：“gave the X to Y”）来探索这一问题，采用受控养育范式，迭代地在系统化操纵的输入上训练小型LMs。我们关注影响交替选项选择的两个属性：长度和生命性。这两个属性直接存在于与格中，但也反映了更全局的倾向，即较短元素先于较长元素，有生命体先于无生命体。首先，通过在输入中操纵和消融这些偏置的与格，我们表明长度和生命性的直接证据很重要，但即使没有此类证据，轻松优先偏好依然存在。然后，使用在系统化扰动数据集上训练的LMs来操纵全局长度效应（在保留依存结构的同时全局重新线性化句子），我们发现与格偏好可以从间接证据中产生。我们得出结论，LMs的涌现句法偏好来自直接和间接来源的混合。
英文摘要：
Language models (LMs) tend to show human-like preferences on a number of syntactic phenomena, but the extent to which these are attributable to direct exposure to the phenomena or more general properties of language is unclear. We explore this with the English dative alternation (DO: "gave Y the X" vs. PO: "gave the X to Y"), using a controlled rearing paradigm wherein we iteratively train small LMs on systematically manipulated input. We focus on two properties that affect the choice of alternant: length and animacy. Both properties are directly present in datives but also reflect more global tendencies for shorter elements to precede longer ones and animates to precede inanimates. First, by manipulating and ablating datives for these biases in the input, we show that direct evidence of length and animacy matters, but easy-first preferences persist even without such evidence. Then, using LMs trained on systematically perturbed datasets to manipulate global length effects (re-linearizing sentences globally while preserving dependency structure), we find that dative preferences can emerge from indirect evidence. We conclude that LMs' emergent syntactic preferences come from a mix of direct and indirect sources.
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。
