{"slug": "relational-over-regularization-graph-based-ai-generated-text-detection-via", "title": "Relational Over-Regularization: Graph-Based AI-Generated Text Detection via Sentence Transition Deviation", "summary": "Researchers propose Relational Over-Regularization (ROR), a sentence-pair-level signal showing LLMs produce inter-sentence transition variance that deviates from human writing, and validate it across four benchmarks (p < 0.001). Their Cross-Source Stylometric Fingerprint Graph (CSFG) achieves 97.14% accuracy in binary AI-generated text detection, outperforming the strongest graph-based baseline by 11.14 percentage points with a 1.57% false-positive rate, though detection degrades for generators with transition variance at or below the human baseline.", "body_md": "arXiv:2608.26694v1 Announce Type: new\nAbstract: Detecting AI-generated text (AIGT) remains challenging because existing approaches rely on token-level statistical signals or independent stylometric features, causing them to overfit to specific generators and fail under distribution shift. We identify a structural signal at the sentence-pair level: LLMs produce inter-sentence transition variance that deviates from human writing through inflated variance driven by recurring similarity bursts at paragraph boundaries and templated transitions. We formalize this as Relational Over-Regularization (ROR) and validate it across four benchmarks (p < 0.001). The central contribution is this relational problem formulation, not a novel GNN architecture; CSFG is one concrete instantiation for operationalizing ROR. To exploit this signal, we propose the Cross-Source Stylometric Fingerprint Graph (CSFG), a graph-based framework that encodes positional, sequential, semantic, and transition deviation signals as learnable GNN edge features. The per-edge signed deviation {\\delta}_ij operationalizes ROR without hand-crafted thresholds and acts as a false-positive calibrator. CSFG achieves 97.14% accuracy under binary detection, outperforming the strongest graph-based baseline by 11.14 pp, with a false-positive rate of 1.57% and robust generalization to unseen LLMs in the inflated-variance regime; detection degrades for generators whose transition variance falls at or below the human baseline.", "url": "https://wpnews.pro/news/relational-over-regularization-graph-based-ai-generated-text-detection-via", "canonical_source": "https://www.machinebrief.com/news/relational-over-regularization-graph-based-ai-generated-text-i4l5", "published_at": "2026-08-28 04:00:00+00:00", "updated_at": "2026-08-28 05:18:58.464282+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "natural-language-processing", "ai-research"], "entities": ["arXiv", "Cross-Source Stylometric Fingerprint Graph (CSFG)", "Relational Over-Regularization (ROR)"], "alternates": {"html": "https://wpnews.pro/news/relational-over-regularization-graph-based-ai-generated-text-detection-via", "markdown": "https://wpnews.pro/news/relational-over-regularization-graph-based-ai-generated-text-detection-via.md", "text": "https://wpnews.pro/news/relational-over-regularization-graph-based-ai-generated-text-detection-via.txt", "jsonld": "https://wpnews.pro/news/relational-over-regularization-graph-based-ai-generated-text-detection-via.jsonld"}}