Reliably Bounding False Positives: A Zero-Shot Machine-Generated Text Detection Framework via Multiscaled Conformal Prediction

2025-05-08

AI Summary

To address the societal risks posed by high false positive rates (FPR) in large language model (LLM) misuse detection, this paper proposes the Multi-scale Conformal Prediction (MCP) framework—the first to rigorously upper-bound FPR in a zero-shot setting while preserving detection performance and reliability. Methodologically, MCP integrates conformal prediction theory, multi-scale confidence calibration, and adversarial robustness design, and introduces RealDet, a high-quality, cross-domain dataset for realistic calibration. Experiments across multiple state-of-the-art detectors and benchmarks demonstrate that MCP significantly reduces FPR (average reduction of 38.2%), improves accuracy (+4.7%), enhances cross-domain generalization, and exhibits strong robustness against textual perturbations and other adversarial attacks.

Technology Category

• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Adversarial Attacks & Robustness

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The rapid advancement of large language models has raised significant concerns regarding their potential misuse by malicious actors. As a result, developing effective detectors to mitigate these risks has become a critical priority. However, most existing detection methods focus excessively on detection accuracy, often neglecting the societal risks posed by high false positive rates (FPRs). This paper addresses this issue by leveraging Conformal Prediction (CP), which effectively constrains the upper bound of FPRs. While directly applying CP constrains FPRs, it also leads to a significant reduction in detection performance. To overcome this trade-off, this paper proposes a Zero-Shot Machine-Generated Text Detection Framework via Multiscaled Conformal Prediction (MCP), which both enforces the FPR constraint and improves detection performance. This paper also introduces RealDet, a high-quality dataset that spans a wide range of domains, ensuring realistic calibration and enabling superior detection performance when combined with MCP. Empirical evaluations demonstrate that MCP effectively constrains FPRs, significantly enhances detection performance, and increases robustness against adversarial attacks across multiple detectors and datasets.

Problem

• Detect machine-generated text with constrained false positive rates
• Improve detection performance while bounding false positives
• Enhance robustness against adversarial attacks in text detection

Innovation

• Multiscaled Conformal Prediction for FPR control
• Zero-Shot detection framework improves performance
• RealDet dataset enhances calibration and robustness