This repository hosts the Human ChatGPT Comparison Corpus (HC3), a foundational dataset for evaluating how closely ChatGPT mimics human responses across various questions. It includes parallel human and ChatGPT answer pairs in both English and Chinese. The project also provides trained detectors that classify text as human-written or AI-generated using the dataset. Published alongside an academic paper, it serves as a resource for studying LLM behavior and building detection systems.