WildWorld: A Large-Scale Dataset for Dynamic World Modeling with Actions and Explicit State toward Generative ARPG

TL;DR

WildWorld is a large-scale, richly annotated dataset from a photorealistic game designed to advance dynamic world modeling with explicit states and actions, addressing limitations of previous datasets.

Contribution

The paper introduces WildWorld, a comprehensive dataset with diverse actions and explicit state annotations, enabling better learning of structured world dynamics in visual environments.

Findings

Models struggle with semantically rich actions.

Maintaining long-horizon state consistency remains challenging.

The dataset facilitates research in state-aware video generation.

Abstract

Dynamical systems theory and reinforcement learning view world evolution as latent-state dynamics driven by actions, with visual observations providing partial information about the state. Recent video world models attempt to learn this action-conditioned dynamics from data. However, existing datasets rarely match the requirement: they typically lack diverse and semantically meaningful action spaces, and actions are directly tied to visual observations rather than mediated by underlying states. As a result, actions are often entangled with pixel-level changes, making it difficult for models to learn structured world dynamics and maintain consistent evolution over long horizons. In this paper, we propose WildWorld, a large-scale action-conditioned world modeling dataset with explicit state annotations, automatically collected from a photorealistic AAA action role-playing game (Monster Hunter: Wilds). WildWorld contains over 108 million frames and features more than 450 actions, including movement, attacks, and skill casting, together with synchronized per-frame annotations of character skeletons, world states, camera poses, and depth maps. We further derive WildBench to evaluate models through Action Following and State Alignment. Extensive experiments reveal persistent challenges in modeling semantically rich actions and maintaining long-horizon state consistency, highlighting the need for state-aware video generation. The project page is https://shandaai.github.io/wildworld-project/.