Translated data: Researchers at the University of California, Berkeley, have introduced an advanced AI method called Ghostbuster for detecting text generated by LLMs. Ghostbuster showcases exceptional performance across various datasets, outperforming other frameworks, thanks to its three-stage training process involving structured search and linear classification. This method addresses the gap in poor performance on non-original evaluation datasets, providing a new tool for assessing the authenticity of text generated by LLMs.