AI文本检测器误判人类正式写作，本质是放大语言模型的预测性轴线。
Amplifying, Not Learning: The Price of Out-of-Distribution Generalization in AI-Text Detection
- 检测器不学习真假边界，而是放大预训练模型固有的可预测性特征
- 人类正式作文被误判为AI生成率高达99.5%，而高温度AI反而仅10.5%
- 该偏差是结构化的，跨语言、跨代码均存在，无法通过调参消除
AI文本检测器在教育、招聘和出版中用于决策，却将最流畅、正式的人类写作误标为机器生成：中位数正式母语人类作文被判定为AI的概率达99.5%，而真实高温度生成内容仅10.5%。部署的检测器（如chatgpt-detector-roberta）在1%假警报率下仍标记56%的正式文章。这并非校准错误，而是单一机制的体现：微调后的检测器并未学习真假边界，而是放大了微调前就存在的典型性轴线（语言模型下的可预测性），仅进行缩放而非构建新边界。分解检测器后发现，继承部分承担了跨生成器迁移的主体，导致对正式人类文本的过度标记；残差部分为生成器特异，不具备迁移性。仅用约25个标签/类的冻结表示探针即达到全微调检测器性能（跨生成器AUROC 0.893 vs 0.831）。这一现象在三种架构、微调解码器和零样本困惑度检测器中一致存在，表明：因可迁移轴线即为过检源头，任何训练目标、阈值、概念擦除或集成方法均无法在保留跨生成器检测能力的同时消除危害。一种闭式、免训练算子可定位并诊断偏差（恢复已失效检测器，真阳性率从0升至0.904，假警报率1%），但跨生成器检测的AUROC保持不变：仅转移偏差，无法根除。该机制不限于英文散文，在中文和代码中同样复现。检测器不公平性是分布外泛化必然带来的结构性代价。
原文摘要 · Abstract (English)
AI-text detectors gate decisions in education, hiring, and publishing, yet they flag the most fluent, formal human writing as machine-generated: they rate the median formal-native human essay as 99.5% likely AI while clearing genuine high-temperature AI at 10.5%. Deployed detectors share it (chatgpt-detector-roberta flags 56% of formal essays at a 1% false-alarm rate). This is not a calibration bug but the signature of one mechanism: a fine-tuned detector does not learn an AI-versus-human boundary, it amplifies an inherited typicality axis (predictability under a language model) that pre-exists fine-tuning, rescaling it rather than constructing one. Decomposing the detector into this inherited reading and a fine-tuned residual, the inherited part carries the bulk of cross-generator transfer and produces the over-flagging of formal humans, while the residual is generator-specific and does not transfer; a frozen-representation probe fit on ~25 labels per class matches the fully fine-tuned detector on unseen generators (cross-generator AUROC 0.893 vs 0.831). This yields a no-go: because the axis that transfers is the axis that over-flags, no training objective, threshold, concept-erasure, or ensemble we test removes the harm while preserving cross-generator detection, across three architectures, fine-tuned decoders, and zero-shot perplexity detectors. A closed-form, training-free operator relocates and diagnoses the bias (reviving a dead deployed detector, true-positive rate 0 to 0.904 at a 1% false-alarm rate) but, consistent with the no-go, is AUROC-neutral on cross-generator detection: it moves the bias, it cannot erase it. The mechanism is not specific to English prose: the over-flag and the decomposition replicate in Chinese and in code. Detector unfairness is a predictable, structural property of the detection paradigm, the price of out-of-distribution generalization.
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。