SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1811.04551 · 2018

Learning Latent Dynamics for Planning from Pixels

Honglak Lee, Ian Fischer, Danijar Hafner, David Ha, Ruben Villegas, Timothy Lillicrap, James Davidson

arXiv · PDF · Open in the Atlas

Code that ran

3 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 0 failures of our instrument; 4 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/1811.04551.json · badge: /api/ran/1811.04551.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
google-research/planet canonical (classifier label)
classifier basis: census
1 ran · 2 no recorded run
cross32768/PlaNet_PyTorch community
Papers with Code archive
1 ran · 0 no recorded run
xingyu-lin/softagent community
Papers with Code archive
1 ran · 0 no recorded run
Kaixhin/PlaNet community
Papers with Code archive
0 ran · 1 no recorded run
Yizhao111/dreamer-pytorch community
Papers with Code archive
0 ran · 1 no recorded run
FunctionStatusWhere it lives
RecurrentStateSpaceModel Ran
ran on a synthesized input · constructed an object; no result was computed
cross32768/PlaNet_PyTorch/model.py
code served (permissive licence) · get_code("5c38c6ed046ae967")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, RecurrentStateSpaceModel: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch import nn
import torch.nn.functional as F

torch.manual_seed(0)

# Define dimensions for different concepts
state_dim = 5
action_dim = 3
rnn_hidden_dim = 7
hidden_dim = 200
embedded_obs_dim = 1024

# Create sample tensors for the forward call
batch_size = 2

# State tensor: (batch_size, state_dim)
state = torch.randn(batch_size, state_dim)

# Action tensor: (batch_size, action_dim)
action = torch.randn(batch_size, action_dim)

# RNN hidden state: (batch_size, rnn_hidden_dim)
rnn_hidden = torch.randn(batch_size, rnn_hidden_dim)

# Embedded observation: (batch_size, embedded_obs_dim)
# Note: the code hardcodes 1024, so we use that
embedded_next_obs = torch.randn(batch_size, 1024)

ARGS = {
    'state_dim': state_dim,
    'action_dim': action_dim,
    'rnn_hidden_dim': rnn_hidden_dim,
    'hidden_dim': hidden_dim,
    'min_stddev': 0.1,
    'act': F.relu,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type RecurrentStateSpaceModel — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a RecurrentStateSpaceModel output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
TransitionModel Ran
ran on a synthesized input · constructed an object; no result was computed
xingyu-lin/softagent/planet/models.py
pointer only (licence: NONE) · get_code("efe7f713034a1846")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, xingyu-lin/softagent when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

lambda_return Ran
ran on a synthesized input
google-research/planet/planet/control/temporal_difference.py
code served (permissive licence) · get_code("e437c521c0aeb696")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned EagerTensor, ndim 3, dtype family float, shape (2, 3, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import tensorflow as tf
import numpy as np

# Set random seed for reproducibility
np.random.seed(0)
tf.random.set_seed(0)

# Create small test tensors
batch_size = 2
time_steps = 3
value_dim = 4

# reward: shape (batch_size, time_steps, value_dim)
reward = tf.constant(np.random.randn(batch_size, time_steps, value_dim), dtype=tf.float32)

# value: shape (batch_size, time_steps, value_dim) - same ndims as reward
value = tf.constant(np.random.randn(batch_size, time_steps, value_dim), dtype=tf.float32)

# bootstrap: shape (batch_size, value_dim) or None
bootstrap = tf.constant(np.random.randn(batch_size, value_dim), dtype=tf.float32)

# discount: scalar float
discount = 0.99

# lambda_: scalar float (note: lambda is reserved in Python, so lambda_)
lambda_ = 0.95

# axis: which axis to use for time dimension
axis = 1

# stop_gradient: boolean
stop_gradient = True

ARGS = {
    "reward": reward,
    "value": value,
    "bootstrap": bootstrap,
    "discount": discount,
    "lambda_": lambda_,
    "axis": axis,
    "stop_gradient": stop_gradient,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 3, dtype family float, shape (2, 3, 4) — shape (2, 3, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
TransitionModel Not yet run
unverified (harvested, no recorded run)
Kaixhin/PlaNet/models.py
code served (permissive licence) · get_code("c2d4d65f835c80d5")
How it ran

no invocation recorded.

TransitionModel Not yet run
unverified (harvested, no recorded run)
Yizhao111/dreamer-pytorch/models.py
code served (permissive licence) · get_code("c2cbf0fc5d0dc1a5")
How it ran

no invocation recorded.

discounted_return Not yet run
unverified (harvested, no recorded run)
google-research/planet/planet/control/temporal_difference.py
code served (permissive licence) · get_code("e0037b2b9ff475af")
How it ran

no invocation recorded.

fixed_step_return Not yet run
unverified (harvested, no recorded run)
google-research/planet/planet/control/temporal_difference.py
code served (permissive licence) · get_code("5ebd8ca0b33c28d7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 1 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 1 later) and 6 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
TransitionModel
brick of this paper's assembly: entry
1 other paper
1 later
1912.01603 (2019-12)
Dream to Control: Learning Behaviors by Latent Imagination

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Planning has been very successful for control tasks with known environment dynamics. To leverage planning in unknown environments, the agent needs to learn the dynamics from interactions with the world. However, learning dynamics models that are accurate enough for planning has been a long-standing challenge, especially in image-based domains. We propose the Deep Planning Network (PlaNet), a purely model-based agent that learns the environment dynamics from images and chooses actions through fast online planning in latent space. To achieve high performance, the dynamics model must accurately predict the rewards ahead for multiple time steps. We approach this using a latent dynamics model with both deterministic and stochastic transition components. Moreover, we propose a multi-step variational inference objective that we name latent overshooting. Using only pixel observations, our agent solves continuous control tasks with contact dynamics, partial observability, and sparse rewards, which exceed the difficulty of tasks that were previously solved by planning with learned models. PlaNet uses substantially fewer episodes and reaches final performance close to and sometimes higher than strong model-free algorithms.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("1811.04551")
get_code_for_paper("1811.04551")
have("1811.04551")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/1811.04551.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/1811.04551.svg)](https://syntology.ai/paper/1811.04551)

Connect an agent — have() is free.