跳到正文
arXiv cs.CL· Xiang Ding·· 11 小时前AI 评分38

语言模型会保留被放弃的分析:花园路径句后的残留与适应

Language Models Keep Abandoned Analyses: Residue and Adaptation after Garden-Path Sentences

AI 导读

研究在 24 个 124M 至 70B 参数的开源模型中发现,花园路径句会在上下文中留下两种痕迹:被放弃分析产生的残留使后续宾语附着倾向提升约四分之三,且该效应出现在采样文本、留出集与人类常模项目上;朝向已消歧结构的适应则随消歧误差增大而增强,主要作用于理解类指标。

正文 · 原文

View PDF HTML (experimental)

Abstract:Reading "While the man hunted the deer ran into the woods" requires giving up the analysis in which "the deer" is the object of "hunted". People who have just read such a sentence still produce more transitive structures, so the abandoned analysis primes later language. We ask whether language models keep using an analysis they have abandoned, and what in the model carries it. In 24 open models from 124M to 70B parameters, a garden-path sentence leaves two traces in the context. A residue of the abandoned analysis raises later object attachment, by about three quarters of the effect of a sentence in which the noun phrase really is an object, and it appears in sampled text, on held-out and human-normed items and for a second type of garden path. An adaptation toward the resolved structure scales with the sentence's disambiguation error and acts mainly on comprehension-like measures. Activation patching separates the two. Later text reads the residue from the role information in the noun phrase's states, which were computed under the object analysis before the disambiguation, and replacing these states removes almost all of the residue in the 32B and 70B models. The adaptation is read from the states that processed the disambiguation. Overridden statements show the same persistence (a corrected value keeps pulling later predictions in 8 of 9 models), and in most models the pull is carried by the positions after the statement. Finally, the human paradigm transfers only with controls for punctuation, because the models also write garden-path sentences themselves, which reverses the paradigm's raw statistic.
Comments: 16 pages, 5 figures, 7 tables
Subjects: Computation and Language (cs.CL)
ACM classes: I.2.7
Cite as: arXiv:2610.36797 [cs.CL]
  (or arXiv:2610.36797v1 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.2610.36797

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Ding Xiang [view email]
[v1] Tue, 27 Oct 2026 07:59:28 UTC (118 KB)

来源:arXiv cs.CL · arxiv.org