The widespread accessibility of Large Language Models (LLMs) to the general public has significantly amplified the dissemination of machine-generated texts (MGTs). Advancements in prompt manipulation have exacerbated the difficulty in discerning the origin of a text (human-authored vs