Research
On the Indistinguishability of Human v/s AI Generated Text
arXiv:2608.26797v1 Announce Type: new Abstract: The rapid improvement of LLMs has made distinguishing AI-generated text from human writing a pressing problem. This challenge is further amplified by pa
arXiv:2608.26797v1 Announce Type: new Abstract: The rapid improvement of LLMs has made distinguishing AI-generated text from human writing a pressing problem. This challenge is further amplified by paraphrasing tools designed to make machine-generated text appear more "human". We study how access to human writing samples can be used to strategically paraphrase machine-generated responses toward the human distribution. Under a multi-sample setting with human and machine responses to the same prompts, we show that repeated paraphrasing moves the machine distribution toward the empirical human distribution under simple mixing and stability conditions. Our results derive an explicit convergence rate, extend the analysis to a finite-sample setting, and characterize how the required number of human samples and paraphrasing rounds scale with the desired error.
Related
- Steer-to-Detect: Probing Hidden Representations for Detection of LLM-Generated Texts
- Efficient Online LLM Watermark Detection via Rao-Blackwellized E-Processes
- Unveiling Spectral Mechanisms in Training-Free LLM Text Detection
- Token-Level Likelihood-Array Regression for Membership Inference and AI-Generated Text Detection
Source: arXiv cs.LG | 2026-08-28