区分内容与表达来源，提升混合生成文本检测精度
D2C-Routing: Dimension-to-Composition Evidence Routing for Mixed-Origin AI-Generated Text Detection
- 分维度路由内容与表达证据，再融合判断生成来源
- 在混合文本数据集上达0.8603的平均真阳性率
- 适合需要精细溯源的AI文本检测场景
AI生成文本检测通常被视为二元文档级判断，但面对内容与表达来源不同的混合生成文本时，该框架失效。本文将检测任务重构为维度到组合的源归属问题，先分别推断内容来源与表达来源，再合成四类协作类型。提出D2C-Routing方法，通过监督维度头路由内容与表达证据，经可学习门控融合层输出最终标签。在基于HART基准重构的MixD2C数据集上，所提系统达到0.8603的四分类平均真阳性率@1%假阳性率，较同分割下RACE-local重跑结果高6.5点。消融实验验证路由设计有效性，错误分析表明区分AI内容+人类表达与全AI生成文本仍是最难边界。代码已开源。
原文摘要 · Abstract (English)
AI-generated text detection is commonly framed as a binary document-level judgment about whether a text is human-written or machine-generated. This framing breaks down for mixed-origin writing, where content origin and expression origin may differ. We cast mixed-origin detection as dimension-to-composition source attribution, inferring content origin and expression origin before composing them into four collaboration types. We propose Dimension-to-Composition Routing (D2C-Routing), which routes content-side and expression-side evidence to supervised dimension heads before a learned gated composition layer predicts the final label. On MixD2C, a reconstructed split derived from the HART mixed-origin benchmark, our disclosed D2C-Routing-based detector system reaches 0.8603 four-way Avg TPR@1%FPR, 6.5 points above the same-split RACE-local rerun. Core ablations support the routing design, while error analysis shows that distinguishing AI-content/human-expression from fully AI-generated text remains the hardest boundary. Code is available at https://github.com/bystander563/d2c-routing-artifact.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。