Abstract
AI-generated text (AIGT) detection can be sensitive to the decoding choices of the source large language model (LLM). We observe that perturbing next-token logits or adjusting sampling temperature can reduce detection performance, providing a clear signal of detector vulnerability to decoding-time distribution changes. Building on this observation, we propose EASE (Entropy-Adaptive Distribution Shaping for Evasion), a training-free and detector-agnostic framework for evading AIGT detectors. EASE computes predictive entropy directly from the source LLM's next-token distribution and uses it to adapt both logit perturbation and sampling temperature, without detector feedback or model fine-tuning. Experiments across three source LLMs and multiple detectors demonstrate consistent reductions in detection performance, with negligible degradation in text quality and negligible inference overhead.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Zhou, J., Wong, K., Wang, J., & Zhou, J. (2026). EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors. https://omanscience.com/en/articles/ease-entropy-adaptive-distribution-shaping-for-evading-ai-generated-text-detectors
MLA 9
Zhou, Jicheng, et al. "EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors." https://omanscience.com/en/articles/ease-entropy-adaptive-distribution-shaping-for-evading-ai-generated-text-detectors.
Chicago (author–date)
Zhou, Jicheng, Kahim Wong, Jialong Wang, and Jiantao Zhou. 2026. "EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors." https://omanscience.com/en/articles/ease-entropy-adaptive-distribution-shaping-for-evading-ai-generated-text-detectors.
Harvard
Zhou, J., Wong, K., Wang, J. and Zhou, J. (2026) 'EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors', Available at: https://omanscience.com/en/articles/ease-entropy-adaptive-distribution-shaping-for-evading-ai-generated-text-detectors.
Vancouver
Zhou J, Wong K, Wang J, Zhou J. EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors. https://omanscience.com/en/articles/ease-entropy-adaptive-distribution-shaping-for-evading-ai-generated-text-detectors
IEEE
J. Zhou, K. Wong, J. Wang, and J. Zhou, "EASE: Entropy-Adaptive Distribution Shaping for Evading AI-generated Text Detectors," https://omanscience.com/en/articles/ease-entropy-adaptive-distribution-shaping-for-evading-ai-generated-text-detectors.