1 paper
François Chollet, Mike Knoop, Gregory Kamradt +1
The ARC-AGI benchmark series serves as a critical measure of few-shot generalization on novel tasks, a core aspect of intelligence. The ARC Prize 2025 global competition targeted t…