黑大盒子：在大型语言模型中追溯形容词顺序偏好
Black Big Boxes: Tracing Adjective Order Preferences in Large Language Models
AI总结：
研究揭示大型语言模型中形容词顺序偏好主要由分布学习决定，但模型能稳健推广至未见组合，且上下文线索影响其输出顺序。
摘要：
在英语和其他语言中，名词短语中的多个形容词遵循复杂的排序模式。这些模式在语言学中被广泛研究，并为评估语言模型（LMs）如何获取渐进性和上下文敏感的词序偏好提供了有用的测试案例。我们试图确定语言模型中的形容词顺序偏好有多大程度可以由分布学习单独解释，以及模型在哪些方面表现出超出表面共现模式的行为。我们发现，LM的预测行为在很大程度上可以由训练数据频率解释：简单的n-gram统计学在很大程度上解释了它们的行为，并且与训练过程中学习的偏好紧密相似。然而，通过分析学习动态，我们发现模型能够稳健地推广到未见过的形容词组合，这表明其行为不能仅归因于对观察到的顺序的记忆。此外，我们展示了LM如何利用句子上下文中的词序线索，通过特征归因方法证明上下文线索是LM输出中形容词顺序的另一个驱动因素。
英文摘要：
In English and other languages, multiple adjectives in noun phrases follow intricate ordering patterns. These patterns have been widely studied in linguistics and provide a useful test case for assessing how language models (LMs) acquire graded and context-sensitive word order preferences. We ask to what extent adjective order preferences in LMs can be explained by distributional learning alone, and where models exhibit behaviour that goes beyond surface co-occurrence patterns. We find that LM predictions are largely explained by training data frequencies: simple n-gram statistics account for much of their behaviour and closely mirror the preferences learned during training. However, by analysing learning dynamics we reveal that models also generalize robustly to unseen adjective combinations, indicating that their behaviour cannot be reduced to memorization of observed orders alone. Moreover, we show how LMs leverage word order cues from sentence context, demonstrating with feature attribution methods that contextual cues are an additional driver of adjective order in LM output.
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。
