AINeutralarXiv – CS AI · Jun 117/10
🧠A large-scale study of 25 million comments from Hacker News and Reddit reveals that accusations of AI-generated content have surged over 1000% since 2023, yet these accusations rarely correlate with actual linguistic markers of AI writing. The research shows that "AI slop" accusations function primarily as social gatekeeping rather than genuine detection, challenging assumptions about how AI impacts online discourse.
AIBearisharXiv – CS AI · May 47/10
🧠Researchers introduce DeGenTWeb, a systematic methodology for identifying websites dominated by LLM-generated content with minimal human input. The study reveals that LLM-dominant sites are significantly more prevalent across the web than previously understood, with detection accuracy declining as LLM capabilities improve, raising questions about content authenticity and search quality.
AINeutralarXiv – CS AI · Jun 106/10
🧠Researchers demonstrate that while machine-text detection evasion attacks can fool standard detectors, stylistic fingerprints of AI-generated content remain detectable through few-shot learning methods. However, a novel paraphrasing approach that mimics human writing styles can evade all current detectors, though multi-document analysis reveals the deception at scale.
AINeutralarXiv – CS AI · Jun 56/10
🧠Researchers demonstrate that large language models can reliably self-recognize their own outputs through implicit signals encoded in generated text, and this capability can be amplified through targeted steering of internal activation patterns. By injecting sparse random vectors into a model's residual stream during generation, they create detectable fingerprints enabling attribution to specific LLMs with over 98% accuracy while maintaining text quality. This approach offers a practical alternative to traditional AI-generated content detection by leveraging models' natural representation structures.
AINeutralarXiv – CS AI · Jun 56/10
🧠Researchers propose RidgeFT, a machine learning framework that enables continuous identification of machine-generated text sources while preserving performance on previously learned generators. The method uses efficient closed-form updates and feature-stable analytics to balance adaptation to new language models with retention of old ones.
AINeutralarXiv – CS AI · Jun 26/10
🧠Researchers propose a distribution-free statistical framework that enhances rewrite-based LLM detection systems with finite-sample false discovery rate (FDR) guarantees without requiring model retraining. By formulating detection as a knockoff-based multiple hypothesis testing problem, the framework enables existing detectors to inherit statistical guarantees through a simple calibration procedure, validated across multiple detection models, domains, and language models.
AINeutralarXiv – CS AI · Jun 16/10
🧠Researchers propose Gap-K%, a novel method for detecting whether text was part of an LLM's pretraining data by analyzing the probability gap between a model's top prediction and the actual target token. The technique outperforms existing approaches on standard benchmarks and addresses critical privacy and copyright concerns surrounding the opaque datasets used to train large language models.
AINeutralarXiv – CS AI · May 276/10
🧠Researchers introduce TADDLE, an AI system that detects quality deficiencies in LLM-generated peer reviews by decomposing analysis into specialized tools and multi-label classification. The work addresses a growing problem in academic publishing where AI-written reviews are fluent but potentially flawed, backed by the first expert-annotated benchmark of 1,800 reviews across six defect categories.
AINeutralarXiv – CS AI · Apr 106/10
🧠Researchers introduce AI-Sinkhole, an AI-agent augmented DNS-blocking framework that dynamically detects and temporarily blocks LLM chatbot services during proctored exams to prevent academic integrity violations. The system uses quantized LLMs for semantic classification and Pi-Hole for network-wide DNS blocking, achieving robust cross-lingual detection with F1-scores exceeding 0.83.
