发表机构
Northeastern University; University at Albany, State University of New York（东北大学; 纽约州立大学奥尔巴尼分校）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究通过掩码引用任务对比6种LLM与人类的引用行为，发现LLM引用批判性更低、过度引用热门旧论文、引用社会距离更远的作者，重塑了科学引用的修辞与范围。
AI 中文摘要
科学引用带有修辞意图，学者可能以积极（支持）、消极（对比）或中立（提及）的方式引用先前研究。随着大语言模型（LLM）越来越多地辅助科学写作，它们是否会复制与人类相同修辞意图的引用仍不明确。我们引入了一个掩码引用任务，以比较人类和LLM生成的引用行为。对于每个引用上下文，LLM生成一个替换引用句子，生成可与人类引用直接比较的反事实语料库。我们分析模型引用的内容、对象和方式，使用LLM作为评判者对引用意图进行分类，并利用一个拥有2000万条边的合著网络来衡量被引用作者之间的社会距离。在6种流行的LLM和1746篇顶级NLP会议论文（6.3万+上下文、13.2万+引用）中，出现了三种模式：（1）与人类引用相比，LLM的批判性明显更低；（2）LLM过度引用受欢迎且较旧的论文，在对比引用中这种倾向被放大，而人类写作更常借鉴近期、小众的研究；（3）人类常引用自己的紧密社交网络，尤其是支持性引用，而LLM倾向于引用社会距离更远的作者。这些差异是双刃剑：LLM的引用范围超出了学者的紧密合作者，同时批判性更低，放大了可见性偏差，重塑了科学引用的修辞与范围。
英文摘要
Scientific citations carry rhetorical intent. Scholars may cite prior work positively (supporting), negatively (contrasting), or neutrally (mentioning). As large language models (LLMs) increasingly assist scientific writing, whether they reproduce citations with the same rhetorical intent as humans remains unclear. We introduce a masked-citation task to compare human and LLM-generated citation behavior. For each citation context, an LLM generates a replacement citation sentence, producing a counterfactual corpus directly comparable to human citation. We analyze what, whom, and how models cite, using an LLM-as-a-judge to classify citation intent and a 20-million-edge coauthorship network to measure social distance between cited authors. Across six popular LLMs and 1,746 top NLP conference papers (63k+ contexts, 132k+ citations), three patterns emerge: (1) Compared with human citation, LLMs cite significantly less critically; (2) LLMs over-cite popular and older papers, a tendency amplified for contrasting citations where human writing more often draws on recent, niche work; (3) Whereas humans often cite within their close social network, especially for supporting citations, LLMs tend to draw on more socially distant authors. Together, these differences are double-edged: LLM citation reaches beyond a scholar's close collaborators while being less critical and amplifying visibility bias, reshaping the rhetoric and reach of scientific citation.
CommentsAccepted at the EMNLP 2026 main conference
