world-models/world-models-explorer

🌍 World Models Explorer

Explore world models for simulation, prediction, planning, agents, robotics and Physical AI.

World Models Explorer is a discovery interface for the rapidly growing world-model ecosystem on Hugging Face.

It is designed for a broad AI audience: developers, researchers, students, robotics engineers, agent builders and anyone trying to understand how world models connect perception, prediction, simulation, planning and action.

What are World Models?

A world model is an AI model that learns an internal representation of an environment and how that environment changes.

A world model can help an AI system answer questions such as:

• —What is happening now?
• —What is likely to happen next?
• —What could happen if an action is taken?
• —Which future states are possible?
• —Which action may lead toward a goal?

A useful mental model is:

Observation
    ↓
World Model
    ↓
Possible Futures
    ↓
Planning
    ↓
Action

World models can operate in pixels, latent representations, 3D environments, multimodal spaces, robot states, simulations or other structured representations.

Explore the ecosystem

This Space helps discover models and projects related to:

• —Latent World Models
• —Predictive World Models
• —Generative World Models
• —Interactive World Models
• —World Foundation Models
• —World Action Models
• —Embodied World Models
• —Robotics World Models
• —Spatial World Models
• —Video World Models
• —Agent World Models
• —Model-Based Reinforcement Learning
• —Physical AI
• —Autonomous Systems

Discovery, not ranking

The Explorer is intentionally not a leaderboard.

Downloads and likes can be useful discovery signals, but they do not determine model quality.

Before using a model, inspect its model card and evaluate:

• —architecture
• —modalities
• —context or horizon
• —action conditioning
• —domain
• —license
• —model size
• —latency
• —compute requirements
• —benchmark methodology
• —training data
• —limitations

World Models Ecosystem

WORLD MODELS
│
├── Representation
├── Dynamics
├── Prediction
├── Simulation
├── Planning
└── Action
    │
    ├── Agents
    ├── Robotics
    ├── Physical AI
    ├── Autonomous Systems
    ├── Reinforcement Learning
    └── Interactive Environments

Related Spaces

The World Models organization is building a structured ecosystem:

• —World Models Explorer — discover the field
• —World Model Registry — structured model metadata
• —World Model Benchmark — evaluation and benchmarks
• —World Model Landscape — visual taxonomy and ecosystem map

Cooperation

We welcome cooperation around:

• —world models
• —model discovery
• —research
• —benchmarks
• —datasets
• —robotics
• —embodied AI
• —Physical AI
• —agent systems
• —simulation
• —open-source tooling
• —ecosystem projects

Cooperation, research and ecosystem partnerships: agenten@magenta.de

World Models Explorer Understand the field. Discover the models. Explore possible futures.