超越原始检测分数：基于马尔可夫的校准以提升机器生成文本检测
Beyond Raw Detection Scores: Markov-Informed Calibration for Boosting Machine-Generated Text Detection
AI总结：
本文提出一种基于马尔可夫随机场的校准策略，用于提升机器生成文本检测的准确性，通过减少token级检测分数的偏差，提高检测性能。
摘要：
尽管机器生成文本（MGT）提供了极大的便利，但它们也带来了诸如虚假信息和钓鱼的风险，突显了可靠检测的必要性。基于指标的方法，通过提取统计上可区分的MGT特征，通常比复杂的模型方法更实用，后者容易过拟合。鉴于其多样的设计，我们首先将具有代表性的基于指标的方法置于统一的框架中，从而能够清晰地评估其优势和局限性。我们的分析发现这些方法面临的核心挑战是，token级别的检测分数容易受到MGT生成过程内在随机性的偏倚。为了解决这一问题，我们理论和实证地揭示了两个可能有助于校准的上下文检测分数关系：邻居相似性和初始不稳定性。然后，我们提出了一种基于马尔可夫的分数校准策略，利用马尔可夫随机场建模这些关系，并通过均场近似实现为一个轻量组件，使我们的方法能够无缝集成到现有检测器中。在各种现实场景中的广泛实验，如跨LLM和改写攻击，证明了与基线相比显著的提升，且计算开销极小。代码可在https://github.com/tmlr-group/MRF_Calibration获取。
英文摘要：
While machine-generated texts (MGTs) offer great convenience, they also pose risks such as disinformation and phishing, highlighting the need for reliable detection. Metric-based methods, which extract statistically distinguishable features of MGTs, are often more practical than complex model-based methods that are prone to overfitting. Given their diverse designs, we first place representative metric-based methods within a unified framework, enabling a clear assessment of their advantages and limitations. Our analysis identifies a core challenge across these methods: the token-level detection score is easily biased by the inherent randomness of the MGTs generation process. To address this, we theoretically and empirically reveal two relationships of context detection scores that may aid calibration: Neighbor Similarity and Initial Instability. We then propose a Markov-informed score calibration strategy that models these relationships using Markov random fields, and implements it as a lightweight component via a mean-field approximation, allowing our method to be seamlessly integrated into existing detectors. Extensive experiments in various real-world scenarios, such as cross-LLM and paraphrasing attacks, demonstrate significant gains over baselines with negligible computational overhead. The code is available at https://github.com/tmlr-group/MRF_Calibration.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
