Triospect: A Three-Dimensional Framework for Robust Statistical AI-Generated Text Detection Against Diverse Attacks
Summary
Proposes Triospect, a three-dimensional framework that enhances AI-generated text detection robustness against 17 types of attacks, achieving 22.3% AUROC improvement over baselines.
View Cached Full Text
Cached at: 07/01/26, 05:32 AM
# Triospect: A Three-Dimensional Framework for Robust Statistical AI-Generated Text Detection Against Diverse Attacks Source: [https://arxiv.org/abs/2606.31074](https://arxiv.org/abs/2606.31074) [View PDF](https://arxiv.org/pdf/2606.31074) > Abstract:Existing AI\-generated text detectors are vulnerable to attacks that manipulate textual characteristics\. In this study, we propose a novel Triospect Detection Framework by using additional perspectives of content \(core ideas\) and expression \(stylistic elements\) within a given text\. Experiments on two benchmarks involving 17 attacks, 12 domains, and 17 source models demonstrate that Triospect is robust against these attacks\. It improves the strong baseline by a significant margin of 22\.3% \(AUROC\) and 13% \(TPR01\) on the Humanize\-16K after\-attack subset, and by 9\.1% \(AUROC\) and 22% \(TPR01\) on the adversarial RAID\. This framework marks a pioneering effort in statistical methods to enhance detection reliability against attacks\. We release our data and code at[this https URL](https://github.com/baoguangsheng/triospect)\. ## Submission history From: Guangsheng Bao \[[view email](https://arxiv.org/show-email/4faa88a3/2606.31074)\] **\[v1\]**Tue, 30 Jun 2026 03:02:42 UTC \(274 KB\)
Similar Articles
Paraphrasing Attack Resilience of Various AI-Generated Text Detection Methods
This paper investigates the resilience of AI-generated text detection methods (fine-tuned RoBERTa, Binoculars, text feature analysis, and ensembles) against paraphrasing attacks, finding that Binoculars-inclusive ensembles are most effective but also most vulnerable to attacks, highlighting a dichotomy between performance and resilience.
Team DACTYL at PAN 2026: Bayesian Data Mixing and Empirical X-risk Minimization for AI-text Detection
This paper presents methods for detecting AI-generated text using Bayesian data mixing and empirical X-risk minimization, achieving high performance on OOD detection with ModernBERT-large and MCGrad classifiers.
Show, Don't TELL: Explainable AI-Generated Text Detection
Introduces TELL, an AI-generated text detection system that provides explainable annotations alongside numerical scores, achieving competitive AUROC of 0.927 while enabling users to judge authorship based on highlighted textual indicators.
TRIDENT: Enhancing Large Language Model Safety with Tri-Dimensional Diversified Red-Teaming Data Synthesis
TRIDENT is a novel framework and dataset synthesis pipeline for enhancing LLM safety through tri-dimensional red-teaming data that covers lexical diversity, malicious intent, and jailbreak tactics. Fine-tuning Llama-3.1-8B on TRIDENT-Edge achieves 14.29% reduction in Harm Score and 20% decrease in Attack Success Rate compared to baseline models.
Pangram 4 Technical Report
Pangram Labs presents Pangram 4, a state-of-the-art AI text detection model achieving 0.9916 AUROC with very low false positive and negative rates, and improved robustness to adversarial attacks and mixed authorship detection.