Google DeepMind 与 UPenn 提出可解释的 AI 文本检测方法
Google DeepMind 与 UPenn 提出一种可解释的 AI 文本检测方法，通过稀疏自编码器提取并自动标注检测器残差流中的特征，再仅用通过保真度检查的特征训练小型分类器。该方法在分布外文本上的表现超过原神经检测器和风格基线，论文已被 EMNLP 2026 接收，代码与模型已发布，论文地址为 arxiv.org/abs/2610.30006。
Brilliant paper on interpretable AI-text detection from Google DeepMind and UPenn.
The strongest AI-text detectors are fine-tuned classifiers that give no reason for a flag, and the readable stylometric alternatives are much less accurate. That gap matters when a flag affects a peer review, an admission, or a student.
This work trains sparse autoencoders on a detector's residual stream, labels each feature automatically, and keeps only features whose descriptions pass a fidelity check. A small classifier is then trained on those features alone.
The kept features correspond to specific linguistic markers of AI-generated text. The interpretable classifier also outperforms both the original neural detector and the stylometric baselines on out-of-distribution text.
Accepted at EMNLP 2026, with code and models released.
arxiv.org/abs/2610.30006
Chat with Paper: academy.dair.ai/papers/build…
来源：dair_ai · x.com