Through multi-agent competition, the simple objective of hide-and-seek, and standard
reinforcement learning algorithms at scale, we find that agents create a self-supervised …

Understanding and reasoning about physics is an important ability of intelligent agents. We
develop the PHYRE benchmark for physical reasoning that contains a set of simple classical …

Tools serve as pivotal interfaces that enable humans to understand and reshape the
environment. With the advent of foundation models, AI systems can utilize tools to expand …

Humans can reason about intuitive physics in fully or partially observed environments even
after being exposed to a very limited set of observations. This sample-efficient intuitive …

A fundamental trait of intelligence is the ability to achieve goals in the face of novel
circumstances, such as making decisions from new action choices. However, standard …

Abstract Creative Problem Solving (CPS) is a sub-area within Artificial Intelligence (AI) that
focuses on methods for solving off-nominal, or anomalous problems in autonomous …

Humans use prior knowledge to efficiently solve novel tasks, but how they structure past
knowledge during learning to enable such fast generalization is not well understood. We …

Humans have an inherent ability to learn novel concepts from only a few samples and
generalize these concepts to different situations. Even though today's machine learning …

Physical reasoning requires forward prediction: the ability to forecast what will happen next
given some initial world state. We study the performance of state-of-the-art forward …

Our ability to plan and build a wide array of physicalstructures, from sand castles to
skyscrapers, is a definingfeature of modern human intelligence. What cognitive toolsenable …