SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2004.07219 · 2020

D4RL: Datasets for Deep Data-Driven Reinforcement Learning

U Berkeley, Sergey Levine, Ofir Nachum, Aviral Kumar, Justin Fu, George Tucker

arXiv · PDF · Open in the Atlas

Code that ran

6 of 12 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 4 honoured the contract we drafted, 0 violated it and 1 has no contract verdict; 1 failure of our instrument; 6 no recorded run), measured 2026-11-21; 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/2004.07219.json · badge: /api/ran/2004.07219.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
rail-berkeley/offline_rl canonical (classifier label)
classifier basis not recorded on the link
1 ran · 0 no recorded run
farama-foundation/d4rl canonical (classifier label)
classifier basis: classifier_2025
0 ran · 3 no recorded run
anuragajay/d4rl community (classifier label)
classifier basis: plan_row
3 ran · 3 no recorded run
kpertsch/d4rl community (classifier label)
classifier basis: plan_row
2 ran · 0 no recorded run
FunctionStatusWhere it lives
clamp Ran
ran, honoured the contract we drafted
anuragajay/d4rl/d4rl/carla/carla_env.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from orybkin/lexa-benchmark, is pointer only (licence: NONE) · get_code("ccc25780948874cb")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from orybkin/lexa-benchmark, and GitHub detected no licence in that repository 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.

get_keys Ran
ran; our contract draft was wrong, not the code
rail-berkeley/offline_rl/d4rl/offline_env.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from seohongpark/HILP, is pointer only (licence: NONE) · get_code("395635db622171fa")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from seohongpark/HILP, and GitHub detected no licence in that repository 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.

get_returns Ran
ran, honoured the contract we drafted
kpertsch/d4rl/d4rl/ope.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from orybkin/lexa-benchmark, is pointer only (licence: NONE) · get_code("0817d0a04cd3f2f1")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from orybkin/lexa-benchmark, and GitHub detected no licence in that repository 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.

get_turn Ran
ran on a synthesized input
anuragajay/d4rl/d4rl/gym_minigrid/fourroom_controller.py
code served (permissive licence) · get_code("79dface7ed4101c8")
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 int, ndim 0, dtype family int, 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 torch

torch.manual_seed(0)

RIGHT = 1
LEFT = 0
TURN_DIRS = [[None, RIGHT, RIGHT, LEFT], [LEFT, None, RIGHT, RIGHT], [RIGHT, LEFT, None, RIGHT], [RIGHT, RIGHT, LEFT, None]]

def get_turn(ori, tgt_ori):
    return TURN_DIRS[ori][tgt_ori]

# The function takes two parameters: ori and tgt_ori
# Both are used as indices into TURN_DIRS
# They should be integers in range [0, 3]

ARGS = {
    "ori": 0,
    "tgt_ori": 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 int, ndim 0, dtype family int, 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
parse_maze Ran
ran, honoured the contract we drafted
anuragajay/d4rl/d4rl/pointmaze/maze_model.py
code served (permissive licence) · get_code("ea95f69594dd2e9c")
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 2, dtype family int, 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 numpy as np

WALL = 10
EMPTY = 11
GOAL = 12

def parse_maze(maze_str):
    lines = maze_str.strip().split('\\')
    width, height = (len(lines), len(lines[0]))
    maze_arr = np.zeros((width, height), dtype=np.int32)
    for w in range(width):
        for h in range(height):
            tile = lines[w][h]
            if tile == '#':
                maze_arr[w][h] = WALL
            elif tile == 'G':
                maze_arr[w][h] = GOAL
            elif tile == ' ' or tile == 'O' or tile == '0':
                maze_arr[w][h] = EMPTY
            else:
                raise ValueError('Unknown tile type: %s' % tile)
    return maze_arr

maze_str = "# G \\ O 0 #"

ARGS = {
    "maze_str": maze_str
}
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 ndarray, ndim 2, dtype family int, 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
ranking_correlation_metric Ran
ran, honoured the contract we drafted
kpertsch/d4rl/d4rl/ope.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from orybkin/lexa-benchmark, is pointer only (licence: NONE) · get_code("ec64670acc5c32df")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from orybkin/lexa-benchmark, and GitHub detected no licence in that repository 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.

compute_distance Not yet run
unverified (harvested, no recorded run)
anuragajay/d4rl/d4rl/carla/carla_env.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from orybkin/lexa-benchmark, is pointer only (licence: NONE) · get_code("b7eec81dfe2b1b23")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from orybkin/lexa-benchmark, and GitHub detected no licence in that repository 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. No invocation recorded.

compute_distance Not yet run
unverified (harvested, no recorded run)
anuragajay/d4rl/d4rl/carla/data_collection_agent_lane.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from orybkin/lexa-benchmark, is pointer only (licence: NONE) · get_code("b5b4d8429fac0ada")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from orybkin/lexa-benchmark, and GitHub detected no licence in that repository 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. No invocation recorded.

download_dataset_from_url Not yet run
unverified (harvested, no recorded run)
farama-foundation/d4rl/d4rl/offline_env.py
code served (permissive licence) · get_code("88d6d21f699682b1")
How it ran

no invocation recorded.

filepath_from_url Not yet run
unverified (harvested, no recorded run)
farama-foundation/d4rl/d4rl/offline_env.py
code served (permissive licence) · get_code("e14321aa0b5c0b78")
How it ran

no invocation recorded.

is_within_distance Not yet run
unverified (harvested, no recorded run)
anuragajay/d4rl/d4rl/carla/carla_env.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from orybkin/lexa-benchmark, is pointer only (licence: NONE) · get_code("de9c8fb5e38fa8ed")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from orybkin/lexa-benchmark, and GitHub detected no licence in that repository 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. No invocation recorded.

load_policy Not yet run
unverified (harvested, no recorded run)
farama-foundation/d4rl/scripts/generation/generate_ant_maze_datasets.py
code served (permissive licence) · get_code("3cf6d03023b905d7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 12 samples lifted from this paper's linked repositories, 12 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 9 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 9 later) and 2 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
clamp 6 other papers
3 dated earlier, 3 later
1911.04175 (2019-11)
Multi-Agent Connected Autonomous Driving using Deep Reinforcement Learning
1911.10868 (2019-11)
End-to-End Model-Free Reinforcement Learning for Urban Driving using Implicit Affordances
2002.00434 (2020-02)
Integrating Deep Reinforcement Learning with Model-based Path Planners for Automated Driving
and 3 more not named here
filepath_from_url 6 other papers
6 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2306.09303 (2023-06)
Datasets and Benchmarks for Offline Safe Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
and 3 more not named here
get_keys 6 other papers
6 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2306.09303 (2023-06)
Datasets and Benchmarks for Offline Safe Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
and 3 more not named here
download_dataset_from_url 4 other papers
4 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
2402.15567 (2024-02)
Foundation Policies with Hilbert Representations
and 1 more not named here
is_within_distance 3 other papers
3 later
ijcai2022_0528 (2022 (year only))
Multi-Constraint Deep Reinforcement Learning for Smooth Action Control
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
parse_maze 3 other papers
3 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
2606.15359 (2026-06)
DiRecT: Safe Diffusion-Based Planning via Receding-Horizon Denoising
compute_distance 2 other papers
2 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
get_returns 2 other papers
2 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play
ranking_correlation_metric 2 other papers
2 later
2211.09817 (2022-11)
On the Effect of Pre-training for Transformer in Different Modality on Offline Reinforcement Learning
2312.01473 (2023-12)
Regularity as Intrinsic Reward for Free Play

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 2026-11-21.

Repositories linked to this paper

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

Abstract

The offline reinforcement learning (RL) setting (also known as full batch RL), where a policy is learned from a static dataset, is compelling as progress enables RL methods to take advantage of large, previously-collected datasets, much like how the rise of large datasets has fueled results in supervised learning. However, existing online RL benchmarks are not tailored towards the offline setting and existing offline RL benchmarks are restricted to data generated by partially-trained agents, making progress in offline RL difficult to measure. In this work, we introduce benchmarks specifically designed for the offline setting, guided by key properties of datasets relevant to real-world applications of offline RL. With a focus on dataset collection, examples of such properties include: datasets generated via hand-designed controllers and human demonstrators, multitask datasets where an agent performs different tasks in the same environment, and datasets collected with mixtures of policies. By moving beyond simple benchmark tasks and data collected by partially-trained RL agents, we reveal important and unappreciated deficiencies of existing algorithms. To facilitate research, we have released our benchmark tasks and datasets with a comprehensive evaluation of existing algorithms, an evaluation protocol, and open-source examples. This serves as a common starting point for the community to identify shortcomings in existing offline RL methods and a collaborative route for progress in this emerging area.

For agents

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

get_harvested_code_for_paper("2004.07219")
get_code_for_paper("2004.07219")
have("2004.07219")

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

curl https://syntology.ai/api/ran/2004.07219.json

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

[![Syntology run record](https://syntology.ai/api/ran/2004.07219.svg)](https://syntology.ai/paper/2004.07219)

Connect an agent — have() is free.