“The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI), introduced in 2019, is a benchmark for evaluating the general fluid intelligence of artificial systems.”