发表机构
Chung-Ang University（中央大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出关系过正则化（ROR），构建基于图的CSFG框架检测AI生成文本，在四个基准上验证，准确率达97.14%，泛化性能优异。
AI 中文摘要
检测AI生成文本（AIGT）仍具挑战性，因为现有方法依赖词元级统计信号或独立文体特征，导致它们对特定生成器过拟合，且在分布偏移下失效。我们识别出句子对层面的结构信号：大型语言模型（LLM）产生的句间转移方差因段落边界处反复出现的相似性爆发和模板化转移而膨胀，偏离人类写作。我们将此形式化为关系过正则化（ROR），并在四个基准上验证（p<0.001）。核心贡献是这种关系问题形式化，而非新型图神经网络（GNN）架构；跨源文体指纹图（CSFG）是实现ROR的具体实例。为利用该信号，我们提出CSFG，这是一种基于图的框架，将位置、序列、语义和转移偏差信号编码为可学习的GNN边特征。每条边的符号偏差δ_ij实现了ROR，无需人工设定阈值，且作为误报校准器。CSFG在二元检测下达到97.14%的准确率，比最强的基于图的基线高出11.14个百分点，误报率为1.57%，且在方差膨胀区域对未见过的LLM具有稳健泛化能力；当生成器的转移方差降至或低于人类基线时，检测性能会下降。
英文摘要
Detecting AI-generated text (AIGT) remains challenging because existing approaches rely on token-level statistical signals or independent stylometric features, causing them to overfit to specific generators and fail under distribution shift. We identify a structural signal at the sentence-pair level: LLMs produce inter-sentence transition variance that deviates from human writing through inflated variance driven by recurring similarity bursts at paragraph boundaries and templated transitions. We formalize this as Relational Over-Regularization (ROR) and validate it across four benchmarks (p < 0.001). The central contribution is this relational problem formulation, not a novel GNN architecture; CSFG is one concrete instantiation for operationalizing ROR. To exploit this signal, we propose the Cross-Source Stylometric Fingerprint Graph (CSFG), a graph-based framework that encodes positional, sequential, semantic, and transition deviation signals as learnable GNN edge features. The per-edge signed deviation δ_ij operationalizes ROR without hand-crafted thresholds and acts as a false-positive calibrator. CSFG achieves 97.14% accuracy under binary detection, outperforming the strongest graph-based baseline by 11.14 pp, with a false-positive rate of 1.57% and robust generalization to unseen LLMs in the inflated-variance regime; detection degrades for generators whose transition variance falls at or below the human baseline.
CommentsPresented in EMNLP 2026
