NVIDIA AI has launched the Nemotron-Personas-Korea dataset, marking the first Korean-language persona dataset. Announced on April 25, the dataset has quickly become one of the top five most popular on the Hugging Face platform. Built on 62 census variables, it supports the largest persona generation model capable of simulating up to 10^46 states, enhancing the reflection of real-world Korean contexts.