New: the Timeline is live. Track world model releases, papers, and benchmark updates in real time.
world-models.io
The Knowledge Hub for AI World Models

Ha & Schmidhuber World Model vs DreamerV3

The original 2018 'World Models' paper vs. the current state-of-the-art: how five years of research transformed a foundational concept into a domain-general world model agent.

robotics model-based-rl simulation embodied-ai

Comparison Overview

Main comparison summary preserved directly in static HTML.

The original 2018 'World Models' paper vs. the current state-of-the-art: how five years of research transformed a foundational concept into a domain-general world model agent.

Verdict

Primary editorial conclusion preserved for non-JS crawlers and readers.

Ha & Schmidhuber's World Model is where it all started: the paper that popularized the idea of training in imagination. DreamerV3 is where it is now: a mature system that masters diverse domains with fixed hyperparameters. Studying both provides the clearest picture of how the field evolved from elegant concept to practical mastery.

Key Differences

Extractable difference list generated from the comparison table.

  • Architecture: Ha & Schmidhuber World Model - VAE + MDN-RNN + Controller; DreamerV3 - RSSM with actor-critic.
  • Latent Space: Ha & Schmidhuber World Model - Continuous VAE; DreamerV3 - Discrete categorical.
  • Domains: Ha & Schmidhuber World Model - Car Racing, VizDoom; DreamerV3 - Atari, DMControl, Minecraft, etc..
  • Policy: Ha & Schmidhuber World Model - Linear controller (CMA-ES); DreamerV3 - Neural actor-critic.
  • Sample Efficiency: Ha & Schmidhuber World Model - Moderate; DreamerV3 - Very high.

When To Use Each

Static decision guidance for no-JS readers.

Choose Ha & Schmidhuber World Model when...

Choose Ha & Schmidhuber World Model when its capabilities best match your research or deployment requirements.

Choose DreamerV3 when...

Choose DreamerV3 when its capabilities best match your research or deployment requirements.

Comparison Table

Ha & Schmidhuber's World Model is where it all started: the paper that popularized the idea of training in imagination. DreamerV3 is where it is now: a mature system that masters diverse domains with fixed hyperparameters. Studying both provides the clearest picture of how the field evolved from elegant concept to practical mastery.

DimensionHa & Schmidhuber World ModelDreamerV3
ArchitectureVAE + MDN-RNN + ControllerRSSM with actor-critic
Latent SpaceContinuous VAEDiscrete categorical
DomainsCar Racing, VizDoomAtari, DMControl, Minecraft, etc.
PolicyLinear controller (CMA-ES)Neural actor-critic
Sample EfficiencyModerateVery high
Historical RoleSeminal: coined 'World Models'Current SOTA for general model-based RL
Year20182023

Performance Index Snapshot

High-level scoring context for the models referenced in this comparison.

ModelCategoryIndex v1.1Confidence
World Models (Ha & Schmidhuber)Model-Based RL48/100high
DreamerV3Model-Based RL88/100high
PlaNetModel-Based RL57/100high
DreamerV2Model-Based RL72/100high

Frequently Asked Questions

FAQ answers rendered directly into static HTML for extractable responses.

Should I start learning with the 2018 paper?

Absolutely. Ha & Schmidhuber's paper is simpler and more conceptually clear. It provides the best introduction to world model concepts before studying the more complex Dreamer family.

Quick Answer

Short extractable summary preserved directly in static HTML.

  • Ha & Schmidhuber World Model vs DreamerV3: this page compares where each system is stronger instead of forcing a universal winner.
  • Use the verdict for the short answer, then validate the trade-offs in the table, evidence sources, and benchmark context.
  • Related models and source links help connect this comparison to the broader world models landscape.

Editorial Trust Signals

Editorial provenance and refresh policy preserved directly in static HTML.

Published by world-models.io editorial board.

Lead editor Bernard Grenat.

This comparison page publishes a direct answer, explicit trade-offs, and source-backed evidence that can be validated against primary materials.

Each editorial page is assembled from primary sources, normalized into extractable summaries, checked for factual drift, and reviewed before publication or major refreshes. Last reviewed: 2026-06-21.

Pages are refreshed when a new paper, benchmark, release, architecture update, or stronger primary source materially changes the answer a reader or AI system should retrieve.

Each page links back to relevant primary sources and keeps a stable canonical URL so readers can verify claims, trace context, and reference the most up-to-date version. See the editorial policy.

Primary sources onlyLast reviewed date visibleMethodology documentedSource links included

External Sources

Primary papers and official sources for the models discussed on this comparison page.