cs.CLNov 11, 2026

FIBER: Bayesian Fusion of Fast-DetectGPT and IRT-Informed Detector Evidence for LLM-Generated Text Detection

Authors: Yan Li, Guangsheng Bao, Zhipeng Chen, Guandong Zhang, Yanbin Zhao

Organizations: School of Mathematics and Statistics, Shanghai Polytechnic University, 2360 Jinhai Road, Pudong New District, 201209, Shanghai, China. · School of Computing and Artificial Intelligence, Fuyao University of Science and Technology, No. 104 Zhihui Avenue, Fuzhou High-tech Industrial Development Zone, 350109, Fujian, China. · School of Computer and Information Engineering, Shanghai Polytechnic University, 2360 Jinhai Road, Pudong New District, 201209, Shanghai, China.

Abstract

Detecting large language model (LLM)-generated text remains challenging under domain and generator shifts. Supervised detectors may generalize poorly, while zero-shot methods often rely on individual statistical signals without leveraging historical detector behavior. We propose FIBER (Fast-DetectGPT with IRT-informed Bayesian Evidence Recalibration), which integrates continuous Fast-DetectGPT scores with auxiliary detector outputs through Bayesian evidence fusion. Item response theory (IRT) models historical detector-correctness patterns, providing class-conditional detector-output likelihoods through marginalization over latent text representations. These likelihoods are combined with Fast-DetectGPT score densities and a historical class prior for prediction. When historical Fast-DetectGPT scores are unavailable, a variant called FIBER-Stream estimates class-specific score distributions from an unlabeled batch. On a 32,000-text multi-domain benchmark, FIBER achieves the highest accuracy (0.620) and improves mean AUC over Fast-DetectGPT (0.664 versus 0.600) under five-fold cross-validation. FIBER also consistently improves AUC across four individual domains. Under out-of-distribution transfer, FIBER-Stream achieved an AUC of 0.911, compared with 0.889 for FIBER and 0.922 for Fast-DetectGPT alone. These results demonstrate the potential of integrating complementary detector evidence and historical information to improve detection performance across heterogeneous settings.

Figures & tables

Explore similar work

CardsList
  1. Detecting the Machine: A Comprehensive Benchmark of AI-Generated Text Detectors Across Architectures, Domains, and Adversarial Conditions

    Mar 18, 2026Madhav S. Baidya, S. S. Baidya, Chirag ChawlaAI-Generated Text DetectionRobustness of Machine-Generated Text Detection

  2. A Measurement Perspective on Benchmarking LLM-Generated Text Detectors using Item Response Theory

    Oct 18, 2026Yanbin Zhao, Zhipeng Chen, Guangsheng Bao +2AI-Generated Text DetectionItem Response Theory

  3. Using LLMs to Detect LLM-Generated Texts: A Cross-Generation Analysis

    Sep 28, 2026Haiyue Yuan, Jie Guo, Weidong Qiu +3LLM EvaluationAI-Generated Text Detection