SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2010.02193 · ICLR · 2021

Mastering Atari with Discrete World Models

Mohammad Norouzi, Jimmy Ba, Danijar Hafner, Timothy Deepmind

arXiv · PDF · Open in the Atlas

Code that ran

11 of 15 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 1 violated it and 9 have no contract verdict; 1 failure 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/2010.02193.json · badge: /api/ran/2010.02193.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
iglu-contest/iglu-builder-baseline-dreamer community (classifier label)
classifier basis not recorded on the link
6 ran · 2 no recorded run
RajGhugare19/dreamerv2 reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 1 no recorded run
andrejorsula/drl_grasping community (classifier label)
classifier basis not recorded on the link
2 ran · 1 no recorded run
FunctionStatusWhere it lives
RSSMUtils Ran
ran on a synthesized input · constructed an object; no result was computed
RajGhugare19/dreamerv2/dreamerv2/models/rssm.py
code served (permissive licence) · get_code("945e8d12ec2f2ec5")
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, RSSMUtils: 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 collections import namedtuple

torch.manual_seed(0)

RSSMContState = namedtuple('RSSMContState', ['mean', 'std', 'stoch', 'deter'])
RSSMDiscState = namedtuple('RSSMDiscState', ['logit', 'stoch', 'deter'])

# Test with continuous RSSM type
rssm_type_cont = 'continuous'
info_cont = {
    'deter_size': 4,
    'stoch_size': 3,
    'min_std': 0.1
}

# Test with discrete RSSM type
rssm_type_disc = 'discrete'
info_disc = {
    'deter_size': 4,
    'class_size': 2,
    'category_size': 3
}

ARGS = {
    'rssm_type': 'continuous',
    'info': {
        'deter_size': 4,
        'stoch_size': 3,
        'min_std': 0.1
    }
}
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 RSSMUtils — 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 RSSMUtils 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
batch_to_seq Ran
ran; our contract draft was wrong, not the code
RajGhugare19/dreamerv2/dreamerv2/models/rssm.py
code served (permissive licence) · get_code("000455412d6f3ed9")
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 Tensor, ndim 4, dtype family float, shape (4, 2, 3, 5) 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 torch

torch.manual_seed(0)

# batch_data: a tensor of shape [L*B, ...] where L*B is the flattened batch dimension
# The function will reshape it to [L, B, ...]
# Let's create a tensor with total batch size = 8 (L*B = 8)
# We'll use L=4, B=2, so L*B = 8
# Add extra dimensions: [8, 3, 5] where 3 and 5 are feature dimensions

batch_data = torch.randn(8, 3, 5)
batch_size = 2
seq_len = 4

ARGS = {
    "batch_data": batch_data,
    "batch_size": batch_size,
    "seq_len": seq_len,
}
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 Tensor, ndim 4, dtype family float, shape (4, 2, 3, 5) — shape (4, 2, 3, 5) 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
flatten_conf Ran
ran on a synthesized input
iglu-contest/iglu-builder-baseline-dreamer/common/other.py
code served (permissive licence) · get_code("db3fffc0e5f25954")
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 dict. 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
import numpy as np

torch.manual_seed(0)

# The function flatten_conf takes a nested dictionary and flattens it
# by concatenating keys with '/' as separator for nested levels.
# It only needs a dictionary as input.

config = {
    'model': {
        'hidden_size': 128,
        'num_layers': 2,
        'dropout': 0.1
    },
    'training': {
        'learning_rate': 0.001,
        'batch_size': 32,
        'epochs': 10
    },
    'data': {
        'train_path': '/path/to/train',
        'val_path': '/path/to/val'
    },
    'optimizer': 'adam'
}

ARGS = {
    'config': config
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
flatten_dict_observations Ran
ran on a synthesized input
andrejorsula/drl_grasping/drl_grasping/utils/utils.py
code served (permissive licence) · get_code("1f9296bbb07fb884")
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 FlattenObservation. 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 gym
import numpy as np

# Create a simple custom environment with Dict observation space
class DictObsEnv(gym.Env):
    def __init__(self):
        self.observation_space = gym.spaces.Dict({
            'position': gym.spaces.Box(low=-1, high=1, shape=(3,), dtype=np.float32),
            'velocity': gym.spaces.Box(low=-1, high=1, shape=(2,), dtype=np.float32),
        })
        self.action_space = gym.spaces.Discrete(4)
    
    def reset(self):
        return {
            'position': np.array([0.0, 0.0, 0.0], dtype=np.float32),
            'velocity': np.array([0.0, 0.0], dtype=np.float32),
        }
    
    def step(self, action):
        obs = {
            'position': np.array([0.1, 0.1, 0.1], dtype=np.float32),
            'velocity': np.array([0.05, 0.05], dtype=np.float32),
        }
        reward = 1.0
        done = False
        info = {}
        return obs, reward, done, info
    
    def render(self, mode='human'):
        pass

env = DictObsEnv()

ARGS = {
    'env': env
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
gym, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type FlattenObservation — a FlattenObservation output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a FlattenObservation 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
load_baselines Ran
ran on a synthesized input
iglu-contest/iglu-builder-baseline-dreamer/common/plot.py
code served (permissive licence) · get_code("b3ae56f77b905c0e")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 collections
import json
import pathlib
import re
import tempfile
from unittest.mock import patch

Run = collections.namedtuple('Run', 'task method seed xs ys')

# Create temporary directory and baseline files for testing
temp_dir = tempfile.TemporaryDirectory()
scores_dir = pathlib.Path(temp_dir.name) / 'scores'
scores_dir.mkdir(parents=True, exist_ok=True)

# Create a sample baselines JSON file
baseline_data = {
    'task1': {
        'baseline_method_a': 0.85,
        'baseline_method_b': 0.90,
        'other_method': 0.75
    },
    'task2': {
        'baseline_method_c': 0.88,
        'custom_method': 0.92
    }
}

baseline_file = scores_dir / 'test_baselines.json'
baseline_file.write_text(json.dumps(baseline_data))

# Create regex patterns to match baseline methods
patterns = [
    re.compile(r'baseline_method_a'),
    re.compile(r'baseline_method_c'),
]

ARGS = {
    'patterns': patterns,
    'prefix': False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
load_episodes Ran
ran on a synthesized input
iglu-contest/iglu-builder-baseline-dreamer/common/replay.py
code served (permissive licence) · get_code("711ef26ab03f56dd")
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 dict. 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 pathlib
import numpy as np
import tempfile
import os

# Create a temporary directory with sample .npz files
temp_dir = tempfile.mkdtemp()

# Create sample episode data
episode1 = {
    'reward': np.array([1.0, 2.0, 3.0, 4.0, 5.0]),
    'observation': np.array([[1, 2], [3, 4], [5, 6], [7, 8], [9, 10]]),
    'action': np.array([0, 1, 0, 1, 0])
}

episode2 = {
    'reward': np.array([2.0, 3.0, 4.0]),
    'observation': np.array([[11, 12], [13, 14], [15, 16]]),
    'action': np.array([1, 0, 1])
}

# Save episodes as .npz files
npz_path1 = os.path.join(temp_dir, 'episode_001.npz')
npz_path2 = os.path.join(temp_dir, 'episode_002.npz')

np.savez(npz_path1, **episode1)
np.savez(npz_path2, **episode2)

ARGS = {
    'directory': temp_dir,
    'limit': None,
    'random': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
lowpass Ran
ran on a synthesized input
andrejorsula/drl_grasping/drl_grasping/utils/wrappers.py
code served (permissive licence) · get_code("aab451ef7c53d13e")
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 ndarray, ndim 1, dtype family float, shape (100,) 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 numpy as np

# Create sample data for filtering
data = np.random.randn(100)  # 1D array of 100 samples

# Set parameters
freq = 5.0  # Corner frequency in Hz
df = 50.0   # Sampling rate in Hz
corners = 4  # Filter order
zerophase = False  # Apply filter once

ARGS = {
    'data': data,
    'freq': freq,
    'df': df,
    'corners': corners,
    'zerophase': zerophase
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (100,) — shape (100,) 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
save_episodes Ran
ran on a synthesized input
iglu-contest/iglu-builder-baseline-dreamer/common/replay.py
code served (permissive licence) · get_code("44b72fcd9872fef3")
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 list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim 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 datetime
import io
import pathlib
import uuid
import numpy as np
import tempfile

# Create a temporary directory for saving episodes
temp_dir = tempfile.mkdtemp()

# Construct sample episodes
# Each episode is a dict with at least 'reward' key (which must be an array)
episodes = [
    {
        'reward': np.array([0.0, 1.0, 2.0, 3.0]),
        'observation': np.array([[1, 2, 3], [4, 5, 6]]),
        'action': np.array([0, 1])
    },
    {
        'reward': np.array([0.5, 1.5]),
        'observation': np.array([[7, 8, 9]]),
        'action': np.array([1])
    }
]

ARGS = {
    'directory': temp_dir,
    'episodes': episodes
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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
schedule Ran
ran on a synthesized input
iglu-contest/iglu-builder-baseline-dreamer/common/other.py
code served (permissive licence) · get_code("c2659ae89b97624c")
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 float, ndim 0, dtype family float, shape () 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

# Test case 1: Simple float string
string_1 = "0.001"
step_1 = tf.constant(100, dtype=tf.int32)

# Test case 2: Linear schedule
string_2 = "linear(0.1,0.01,1000)"
step_2 = tf.constant(500, dtype=tf.int32)

# Test case 3: Warmup schedule
string_3 = "warmup(100,0.1)"
step_3 = tf.constant(50, dtype=tf.int32)

# Test case 4: Exponential schedule
string_4 = "exp(0.1,0.001,500)"
step_4 = tf.constant(250, dtype=tf.int32)

# Test case 5: Horizon schedule
string_5 = "horizon(10,100,1000)"
step_5 = tf.constant(500, dtype=tf.int32)

ARGS = {
    "string": string_1,
    "step": step_1,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () 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
seq_to_batch Ran
ran, violated the contract we drafted
RajGhugare19/dreamerv2/dreamerv2/models/rssm.py
code served (permissive licence) · get_code("7dcf99ab88568073")
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 Tensor, ndim 2, dtype family float, shape (10, 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 torch

torch.manual_seed(0)

# Create a small sequence tensor
# Shape: (sequence_length, batch_size, feature_dim)
# Using distinct sizes: 5 for seq_len, 2 for batch_size, 4 for features
sequence_data = torch.randn(5, 2, 4)

batch_size = 2
seq_len = 5

ARGS = {
    "sequence_data": sequence_data,
    "batch_size": batch_size,
    "seq_len": seq_len,
}
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 Tensor, ndim 2, dtype family float, shape (10, 4) — shape (10, 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
static_scan Ran
ran on a synthesized input
iglu-contest/iglu-builder-baseline-dreamer/common/other.py
code served (permissive licence) · get_code("77ff0f8ba58bad0e")
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 2, dtype family float, shape (2, 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

# Define a simple function that takes (state, input) and returns new state
def simple_fn(state, inp):
    # state and inp are both single tensors
    # This function just adds them together and returns a new state
    return state + inp

# Create small test tensors
# inputs should have shape (sequence_length, feature_dim)
sequence_length = 2
feature_dim = 4
inputs = tf.constant([[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]], dtype=tf.float32)

# start is the initial state, should have same shape as individual outputs
start = tf.constant([0.0, 0.0, 0.0, 0.0], dtype=tf.float32)

ARGS = {
    "fn": simple_fn,
    "inputs": inputs,
    "start": start,
    "reverse": False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 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
RSSM Not yet run
unverified (harvested, no recorded run)
RajGhugare19/dreamerv2/dreamerv2/models/rssm.py
code served (permissive licence) · get_code("11fa7e68493b760d")
How it ran

no invocation recorded.

get_wrapper_class Not yet run
unverified (harvested, no recorded run)
andrejorsula/drl_grasping/drl_grasping/utils/utils.py
code served (permissive licence) · get_code("624c9428ed6a8179")
How it ran

no invocation recorded.

load_run Not yet run
unverified (harvested, no recorded run)
iglu-contest/iglu-builder-baseline-dreamer/common/plot.py
code served (permissive licence) · get_code("f90fbc5c43d007c0")
How it ran

no invocation recorded.

load_runs Not yet run
unverified (harvested, no recorded run)
iglu-contest/iglu-builder-baseline-dreamer/common/plot.py
code served (permissive licence) · get_code("4878c13e37288f19")
How it ran

no invocation recorded.

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 distinct functions with an identity, 6 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 6 later) and 9 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
schedule 2 other papers
2 later
2110.14565 (2021-10)
DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations
2306.09205 (2023-06)
Reward-Free Curricula for Training Robust World Models
static_scan 2 other papers
2 later
2110.14565 (2021-10)
DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations
2306.09205 (2023-06)
Reward-Free Curricula for Training Robust World Models
load_baselines 1 other paper
1 later
2110.14565 (2021-10)
DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations
load_run 1 other paper
1 later
2110.14565 (2021-10)
DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations
load_runs 1 other paper
1 later
2110.14565 (2021-10)
DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations
save_episodes 1 other paper
1 later
2110.14565 (2021-10)
DreamerPro: Reconstruction-Free Model-Based Reinforcement Learning with Prototypical Representations

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

Intelligent agents need to generalize from past experience to achieve goals in complex environments. World models facilitate such generalization and allow learning behaviors from imagined outcomes to increase sample-efficiency. While learning world models from image inputs has recently become feasible for some tasks, modeling Atari games accurately enough to derive successful behaviors has remained an open challenge for many years. We introduce DreamerV2, a reinforcement learning agent that learns behaviors purely from predictions in the compact latent space of a powerful world model. The world model uses discrete representations and is trained separately from the policy. DreamerV2 constitutes the first agent that achieves human-level performance on the Atari benchmark of 55 tasks by learning behaviors inside a separately trained world model. With the same computational budget and wall-clock time, Dreamer V2 reaches 200M frames and surpasses the final performance of the top single-GPU agents IQN and Rainbow. DreamerV2 is also applicable to tasks with continuous actions, where it learns an accurate world model of a complex humanoid robot and solves stand-up and walking from only pixel inputs.

For agents

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

get_harvested_code_for_paper("2010.02193")
get_code_for_paper("2010.02193")
have("2010.02193")

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

curl https://syntology.ai/api/ran/2010.02193.json

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

[![Syntology run record](https://syntology.ai/api/ran/2010.02193.svg)](https://syntology.ai/paper/2010.02193)

Connect an agent — have() is free.