Black-box Detection of LLM-generated Text Using Generalized Jensen-Shannon Divergence

Shuangyi Chen, Ashish J Khisti

Keywords: Text Detection, Likelihood-free Hypothesis Testing, Generalized Jensen-Shannon Divergence

TL;DR: We propose a reference-based detector using surprisal-state Markov transitions and GJS score to flag AI text—fast, accurate, no regeneration.

Abstract: We study black-box detection of machine-generated text under practical constraints: the scoring model (proxy LM) may mismatch the unknown source model, and per-input contrastive generation is costly. We propose SurpMark, a reference-based detector that summarizes a passage by the dynamics of its token surprisals. SurpMark discretizes surprisals into interpretable states, estimates a state-transition matrix for the test text, and scores it via a generalized Jensen–Shannon (GJS) gap between the test transitions and two fixed references (human vs. machine) built once from existing corpora. {\color{blue} Theoretically, we derive design guidance for how the discretization bins should scale with data and provide a principled justification for our test statistic.} Empirically, across multiple datasets, source models, and scenarios, SurpMark consistently matches or surpasses baselines; {\color{blue} our experiments on hyperparameter sensitivity exhibit trends that our theoretical results help to explain, and are consistent with the method’s underlying intuitions.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 7932