---
title: "Predictron | Abstract World Model for Planning | world-models.io"
canonical: "https://world-models.io/en/models/predictron/"
description: "The Predictron learns abstract world models for value prediction and planning in reinforcement learning. Review Predictron: evidence, scope, limitations."
author: "Tyler D."
generated: "2026-10-30"
---

New: the Timeline is live. Track world model releases, papers, and benchmark updates in real time.

The Knowledge Hub for AI World Models

# Predictron

The Predictron is an abstract world model that learns to predict values through multi-step internal simulations, combining model learning and planning end-to-end.

robotics

model-based-rl

simulation

embodied-ai

## What is a world model?

A world model is an AI system that learns a predictive representation of how an environment changes over time. It can use observations, actions, or language to estimate future states and possible outcomes before an agent acts. World models support planning in model-based reinforcement learning, robotics, embodied AI, autonomous driving, and interactive simulation. This site catalogs these systems with model records, research papers, benchmark context, and source links. Start with the [world models database](https://world-models.io/en/world-models-database/), compare systems on the [leaderboard](https://world-models.io/en/leaderboard/), or read the [Performance Index methodology](https://world-models.io/en/methodology/performance-index/).

## Key Attributes

Static HTML snapshot of the model record for crawlers and no-JS readers.

| Attribute | Value |
| --- | --- |
| Model | Predictron |
| Lab / Organization | DeepMind |
| Category | Model-Based RL |
| Subtype | Abstract World Model |
| World Model Type | Abstract internal dynamics model |
| Primary Domain | Value prediction |
| Architecture | Multi-step abstract model with λ-weighted returns |
| Modality | Abstract state representations |
| Training Method | End-to-end supervised learning with multi-step abstract predictions |
| Status | foundational |
| Year | 2017 |
| Performance Index | 43/100 (medium confidence, v1.1) |

## About Predictron

Main editorial body preserved directly in static HTML.

The Predictron combines learning and planning by performing multiple steps of abstract lookahead within a neural network. It learns abstract internal dynamics optimized directly for value prediction, without requiring explicit environment reconstruction. It was one of the earliest demonstrations that world models could be learned end-to-end.

Predictron is an abstract internal dynamics model developed by Google DeepMind in 2017 for value prediction.

## Editorial Snapshot

Short extractable facts for answer engines and no-JS readers.

| Signal | Value |
| --- | --- |
| Definition | Predictron is an abstract internal dynamics model developed by Google DeepMind in 2017 for value prediction. |
| Short Description | An architecture that integrates learning and planning into a single differentiable network via abstract world models. |
| Benchmark Rows | 1 |
| FAQ Entries | 1 |
| Related Models | 2 |
| Related Guides | 0 |
| Related Research Topics | 2 |
| Last Updated | 2026-01-28 |

## Notable Features

Key capabilities associated with this model.

- • End-to-end differentiable planning
- • Abstract dynamics (no reconstruction)
- • λ-weighted returns across imagined depths
- • Pioneering concept in learned planning

## Use Cases

Representative applications attached to this model record.

Value prediction

Abstract planning

Model-based reasoning

## Strengths and Limitations

Balanced assessment surfaced in static HTML.

### Strengths

- • End-to-end differentiable
- • No explicit reconstruction needed
- • Elegant theoretical framework
- • Pioneering concept

### Limitations

- • Limited to value prediction tasks
- • Abstract space not interpretable
- • Superseded by modern approaches

## Benchmarks

Published benchmark evidence attached to this model record.

| Benchmark | Metric | Result | Source |
| --- | --- | --- | --- |
| Grid-world planning | RMSE 0.12 RMSE↓ | Strong improvement over model-free | [Source](https://arxiv.org/abs/1612.08810) |

## References and Citations

Primary references preserved in static HTML for citation extraction.

| Reference | Link |
| --- | --- |
| Silver et al., 2017. The Predictron: End-to-End Learning and Planning. ICML 2017. | [Open source](https://arxiv.org/abs/1612.08810) |

## Related Models

Nearby models linked from the current editorial record.

| Model | Category | World Model Type | Index v1.1 |
| --- | --- | --- | --- |
| [MuZero](https://world-models.io/en/models/muzero/) | Model-Based RL | Abstract learned dynamics + MCTS | 78/100 |
| [PlaNet](https://world-models.io/en/models/planet/) | Model-Based RL | Latent space planning model | 57/100 |

## Direct Comparisons

Side-by-side comparisons already connected to this model.

| Comparison | Matchup | Summary |
| --- | --- | --- |
| [DreamerV3 vs MuZero](https://world-models.io/en/compare/dreamer-v3-vs-muzero/) | DreamerV3 vs MuZero | Both are landmark world model systems, but with fundamentally different architectures. DreamerV3 uses latent imagination with actor-critic learning, while MuZero uses abstract learned dynamics with Monte Carlo tree search. |
| [MuZero vs TD-MPC2](https://world-models.io/en/compare/muzero-vs-td-mpc2/) | MuZero vs TD-MPC2 | Both use learned dynamics models for planning, but MuZero uses Monte Carlo tree search for deep discrete planning while TD-MPC2 uses model-predictive control for continuous multi-task settings. |
| [Predictron vs MuZero](https://world-models.io/en/compare/predictron-vs-muzero/) | Predictron vs MuZero | Both learn abstract dynamics models for planning without requiring environment reconstruction, but Predictron was an early prototype while MuZero became the definitive realization of value-equivalent model learning. |
| [Predictron vs MuZero](https://world-models.io/en/compare/predictron-vs-muzero/) | Predictron vs MuZero | Two DeepMind models that learn abstract value-equivalent dynamics. The Predictron (2017) introduced the concept of learned transition models in abstract space; MuZero (2020) scaled this to superhuman game play without knowing the rules. |

## Research Topics Referencing This Model

Connected research areas surfaced directly in static HTML.

| Topic | Summary |
| --- | --- |
| [Model-Based Reinforcement Learning](https://world-models.io/en/research/model-based-rl/) | What model-based reinforcement learning is, how world models enable imagination-based planning, and why Dreamer, MuZero, PlaNet, and TD-MPC2 matter. |
| [World Models: A Comprehensive Survey](https://world-models.io/en/research/world-models-survey/) | A survey of AI world models covering taxonomy, leading architectures, landmark systems, open challenges, and future research directions. |

## Frequently Asked Questions

FAQ answers rendered directly into static HTML for extractable responses.

### What makes the Predictron different?

Instead of learning an explicit environment model, the Predictron learns abstract internal dynamics optimized directly for value prediction, without needing to reconstruct observations.

## Quick Answer

Short extractable summary preserved directly in static HTML.

- Predictron is an abstract internal dynamics model developed by Google DeepMind in 2017 for value prediction.
- Use this page when you need a fast read on how Predictron fits into the model-based rl landscape, then validate the details in the benchmarks, citations, and related pages.
- A key strength surfaced in the editorial record is end-to-end differentiable.

## Editorial Trust Signals

Editorial provenance and refresh policy preserved directly in static HTML.

**Published by** world-models.io editorial board.

**Lead editor** Tyler D. - Technical editor, methodology and benchmark analysis.

This model page synthesizes primary papers, official model pages, benchmark evidence, and related world-models.io context into a reference resource.

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: 2026-01-28.

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.](https://world-models.io/en/editorial-policy/)

Primary sources only

Last reviewed date visible

Methodology documented

Source links included

Correction channel published

Review date visible

## External Sources

Primary model and lab sources embedded in static HTML.

- [Predictron paper](https://arxiv.org/abs/1612.08810) Google DeepMind primary research source
- [Google DeepMind official website](https://deepmind.google) Related lab

## References

1. [[1] Silver et al., 2017. The Predictron: End-to-End Learning and Planning. ICML 2017.](https://arxiv.org/abs/1612.08810)