Papers › IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures

IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures

5 Feb 2018ICML 2018 7arXiv:1802.01561archive 2025-07-28

Lasse Espeholt, Hubert Soyer, Remi Munos, Karen Simonyan, Volodymir Mnih, Tom Ward, Yotam Doron, Vlad Firoiu, Tim Harley, Iain Dunning, Shane Legg, Koray Kavukcuoglu

In this work we aim to solve a large collection of tasks using a single reinforcement learning agent with a single set of parameters. A key challenge is to handle the increased amount of data and extended training time. We have developed a new distributed agent IMPALA (Importance Weighted Actor-Learner Architecture) that not only uses resources more efficiently in single-machine training but also scales to thousands of machines without sacrificing data efficiency or resource utilisation. We achieve stable learning at high throughput by combining decoupled acting and learning with a novel off-policy correction method called V-trace. We demonstrate the effectiveness of IMPALA for multi-task reinforcement learning on DMLab-30 (a set of 30 tasks from the DeepMind Lab environment (Beattie et al., 2016)) and Atari-57 (all available Atari games in Arcade Learning Environment (Bellemare et al., 2013a)). Our results show that IMPALA is able to achieve better performance than previous agents with less data, and crucially exhibits positive transfer between tasks as a result of its multi-task approach.

PaperPDFConference PDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

The archive's official repository (deepmind/scalable_agent) was not harvested; the 34 samples below come from 11 community repositories.

Syntology 34 code samples harvested from 11 repositories linked to this paper: 16 ran (of which 6 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 18 unverified. Where Syntology's instrument failed: 2 where our contract draft was wrong; 4 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 34 samples from 11 repositories, 16 ran (10 with no instrument failure; 6 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 6 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

24 repositories listed; official and paper-mentioned ones first.

deepmind/scalable_agent officialmentioned in papermentioned on GitHubtfSyntology: not harvested report
Sheepsody/Batched-Impala-PyTorch mentioned on GitHubpytorchSyntology: 2 samples, 1 ran (0 instrument-failed) · 1 unverified report
crazydonkey200/neural-symbolic-machines mentioned on GitHubtfApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Memory Augmented Policy Optimization for Program Synthesis and Semantic Parsing (arXiv:1807.02322; that paper's own run record: 7 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 0 where Syntology's instrument failed) · 8 unverified); nothing checks that they implement this paper's method. report
deepmind/streetlearn mentioned on GitHubtfApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to The StreetLearn Environment and Dataset (arXiv:1903.01292; that paper's own run record: 2 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified); nothing checks that they implement this paper's method. report
facebookresearch/gala mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
facebookresearch/torchbeast mentioned on GitHubpytorchSyntology: 3 samples, 3 ran (3 instrument-failed) · 0 unverified report
google-deepmind/scalable_agent mentioned on GitHubtfSyntology: 3 samples, 1 ran (1 instrument-failed) · 2 unverified report
google-deepmind/streetlearn mentioned on GitHubtfApache-2.0Syntology: not harvested report
google-research/valan mentioned on GitHubtfSyntology: 3 samples, 1 ran (1 instrument-failed) · 2 unverified report
haje01/impala mentioned on GitHubpytorchSyntology: 1 sample, 0 ran · 1 unverified report
heiner/scalable_agent mentioned on GitHubtfApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to TorchBeast: A PyTorch Platform for Distributed RL (arXiv:1910.03552; that paper's own run record: 6 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 4 where Syntology's instrument failed) · 3 unverified); nothing checks that they implement this paper's method. report
jerrodparker20/adaptive-transformers-in-rl mentioned on GitHubpytorchSyntology: not harvested report
michaelnny/deep_rl_zoo mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Playing Atari with Deep Reinforcement Learning (arXiv:1312.5602; that paper's own run record: 64 ran (of which 24 constructed an object rather than computing a result; 46 with no instrument failure: 5 honoured, 0 violated, 41 with no contract checked; 18 where Syntology's instrument failed) · 53 unverified) and Never Give Up: Learning Directed Exploration Strategies (arXiv:2002.06038; that paper's own run record: 7 ran (of which 3 constructed an object rather than computing a result; 5 with no instrument failure: 0 honoured, 0 violated, 5 with no contract checked; 2 where Syntology's instrument failed) · 7 unverified); nothing checks that they implement this paper's method. report
theSparta/neural-symbolic-machines mentioned on GitHubtfApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Memory Augmented Policy Optimization for Program Synthesis and Semantic Parsing (arXiv:1807.02322; that paper's own run record: 7 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 0 where Syntology's instrument failed) · 8 unverified); nothing checks that they implement this paper's method. report
threewisemonkeys-as/torched_impala mentioned on GitHubpytorchSyntology: not harvested report
urw7rs/spiralpp mentioned on GitHubpytorchSyntology: 2 samples, 0 ran · 2 unverified report
villinvic/Georges mentioned on GitHubSyntology: 14 samples, 7 ran (0 instrument-failed) · 7 unverified report
windstrip/DeepMind-StreetLearn mentioned on GitHubtfSyntology: not harvested report
deepmind/haiku jaxSyntology: not harvested report
deepmind/rlax jaxSyntology: 1 sample, 1 ran (1 instrument-failed) · 0 unverified report
opendilab/DI-engine pytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Hindsight Experience Replay (arXiv:1707.01495; that paper's own run record: 16 ran (of which 10 constructed an object rather than computing a result; 11 with no instrument failure: 1 honoured, 0 violated, 10 with no contract checked; 5 where Syntology's instrument failed) · 0 unverified), Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context (arXiv:1901.02860; that paper's own run record: 67 ran (of which 37 constructed an object rather than computing a result; 49 with no instrument failure: 4 honoured, 1 violated, 44 with no contract checked; 18 where Syntology's instrument failed) · 76 unverified), Decision Transformer: Reinforcement Learning via Sequence Modeling (arXiv:2106.01345; that paper's own run record: 17 ran (of which 10 constructed an object rather than computing a result; 12 with no instrument failure: 0 honoured, 0 violated, 12 with no contract checked; 5 where Syntology's instrument failed) · 9 unverified) and 1 more (every one in this page's JSON twin); nothing checks that they implement this paper's method. report
ray-project/ray Syntology: not harvested report
seolhokim/DistributedRL-Pytorch-Ray pytorchSyntology: 2 samples, 1 ran (0 instrument-failed) · 1 unverified report
seolhokim/SimpleDistributedRL pytorchSyntology: 2 samples, 0 ran · 2 unverified report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

34 samples harvested: 16 ran (of which 6 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 1 violated, 8 with no contract checked; 6 where Syntology's instrument failed) · 18 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
1ran · violated contract
2ran · our draft was wrong
4ran; the contract check could not run
8ran
18unverified

Licence: 3 of the 34 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 11 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity.

Sheepsody/Batched-Impala-PyTorch community (archive-listed) 2 samples, 1 ran (0 instrument-failed) · 1 unverified
VTrace src/IMPALA.py ran · metamorphic tier: deterministic no licence file found · pointer only · a7ffabde1fe6a3c9 · report fixture: returned VTrace · level recorded 2026-09-22
Impala src/IMPALA.py unverified no licence file found · pointer only · cd20ae279ac1aa9f · report
facebookresearch/gala community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
GALA_A2C gala/gala_a2c.py ran licence not identified · pointer only · aad00ea22af3c855 · report fixture: returned GALA_A2C · level recorded 2026-09-22
facebookresearch/torchbeast community (archive-listed) 3 samples, 3 ran (3 instrument-failed) · 0 unverified
action_log_probs torchbeast/core/vtrace.py ran · our draft was wrong Apache-2.0 (permissive) · 815b225c5ab9f314 · report fixture: signature policy_logits:Tensor/3/float, actions:Tensor/2/int · returned Tensor · level recorded 2026-09-22
from_importance_weights torchbeast/core/vtrace.py ran · our draft was wrong Apache-2.0 (permissive) · 123172de8a1bc477 · report fixture: signature log_rhos:Tensor/2/float, discounts:Tensor/2/float, rewards:Tensor/2/float, values:Tensor/2/float, bootstrap_value:Tensor/1/float, clip_rho_threshold:float/0/float, clip_pg_rho_threshold:float/0/float · returned VTraceReturns · level recorded 2026-09-22
from_logits torchbeast/core/vtrace.py ran; the contract check could not run (RAISES) Apache-2.0 (permissive) · 715b439b8761ea76 · report fixture: signature behavior_policy_logits:Tensor/3/float, target_policy_logits:Tensor/3/float, actions:Tensor/2/int, discounts:Tensor/2/float, rewards:Tensor/2/float, values:Tensor/2/float, bootstrap_value:Tensor/1/float, clip_rho_threshold:float/0/float, clip_pg_rho_threshold:float/0/float · returned VTraceFromLogitsReturns · level recorded 2026-09-22
google-deepmind/scalable_agent community (archive-listed) 3 samples, 1 ran (1 instrument-failed) · 2 unverified
log_probs_from_logits_and_actions vtrace.py ran; the contract check could not run (RAISES) Apache-2.0 (permissive) · a75de4dc4c61d7dc · report fixture: signature policy_logits:EagerTensor/3/float, actions:EagerTensor/2/int · returned EagerTensor · level recorded 2026-09-22
from_importance_weights vtrace.py unverified Apache-2.0 (permissive) · e11ee068514828e6 · report
from_logits vtrace.py unverified Apache-2.0 (permissive) · 2b5d90526b4ad312 · report
google-research/valan community (archive-listed) 3 samples, 1 ran (1 instrument-failed) · 2 unverified
log_probs_from_logits_and_actions framework/vtrace.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant Apache-2.0 (permissive) · a872fa719c5242ff · report fixture: signature policy_logits:EagerTensor/3/float, actions:EagerTensor/2/int · returned EagerTensor · level recorded 2026-09-22
from_importance_weights framework/vtrace.py unverified Apache-2.0 (permissive) · 86f7ee994e64dc88 · report
from_logits framework/vtrace.py unverified Apache-2.0 (permissive) · c0ca8087e51d7e1d · report
haje01/impala community (archive-listed) 1 sample, 0 ran · 1 unverified
calc_loss_and_backprop learner.py unverified MIT (permissive) · cb89e31851cc2931 · report
urw7rs/spiralpp community (archive-listed) 2 samples, 0 ran · 2 unverified
action_log_probs torchbeast/core/vtrace.py unverified Apache-2.0 (permissive) · bc2ac55256ed974b · report
from_logits torchbeast/core/vtrace.py unverified Apache-2.0 (permissive) · e26d17c678348c08 · report
villinvic/Georges community (archive-listed) 14 samples, 7 ran (0 instrument-failed) · 7 unverified
Categorical in 6 papers training/RL.py ran MIT (permissive) · 89ed626ce6f952a2 · report fixture: signature x:EagerTensor/2/float, param:dict/0/? · returned EagerTensor · level recorded 2026-09-24
CategoricalActor training/RL.py ran · metamorphic tier: deterministic MIT (permissive) · e05eecb1a5750a6c · report fixture: returned CategoricalActor · level recorded 2026-09-22
ConfigLoader training/RL.py ran MIT (permissive) · 4d9828f13202e523 · report fixture: signature config_file:str/0/str · returned ConfigParser · level recorded 2026-09-24
Distribution in 5 papers training/RL.py ran MIT (permissive) · 6490dcf1e1570c31 · report fixture: returned Distribution · level recorded 2026-09-22
V training/RL.py ran · metamorphic tier: deterministic MIT (permissive) · 7d75b7a3afdfddc6 · report fixture: returned V · level recorded 2026-09-22
pairwise_cross_corr training/RL.py ran · honoured contract fingerprinted MIT (permissive) · ba92bb44edd2ada6 · report fixture: signature La:ndarray/2/float, Lb:ndarray/2/float · returned ndarray · level recorded 2026-09-22
safe_crossover training/RL.py ran · violated contract fingerprinted MIT (permissive) · d9e1fb4c37c2e728 · report fixture: signature ax:ndarray/1/float, bx:ndarray/1/float · returned ndarray · level recorded 2026-09-22
AC training/RL.py unverified MIT (permissive) · f34df61f30b311d5 · report
ActionStateProbs training/RL.py unverified MIT (permissive) · 103308756a178abe · report
Default training/RL.py unverified MIT (permissive) · ceddd134b3dd01f7 · report
PlayerState training/RL.py unverified MIT (permissive) · cbb6758024239ec3 · report
corr_matrix training/RL.py unverified MIT (permissive) · 9f5db323eec0eeba · report
layer_crossover training/RL.py unverified MIT (permissive) · a22618d583e622fe · report
nn_crossover training/RL.py unverified MIT (permissive) · e1569fbaca09c124 · report
deepmind/rlax community (archive-listed) 1 sample, 1 ran (1 instrument-failed) · 0 unverified
vtrace rlax/_src/vtrace.py ran; the contract check could not run (RAISES) Apache-2.0 (permissive) · e565e1e08272948c · report fixture: signature v_tm1:ArrayImpl/1/float, v_t:ArrayImpl/1/float, r_t:ArrayImpl/1/float, discount_t:ArrayImpl/1/float, rho_tm1:ArrayImpl/1/float, lambda_:float/0/float, clip_rho_threshold:float/0/float, stop_target_gradients:bool/0/bool, sample_mask:ArrayImpl/1/bool · returned ArrayImpl · level recorded 2026-09-22
seolhokim/DistributedRL-Pytorch-Ray community (archive-listed) 2 samples, 1 ran (0 instrument-failed) · 1 unverified
Learner in 17 papers agents/runners/learners/impala_learner.py ran MIT (permissive) · 3fdd0c84f214c2bb · report fixture: returned Learner · level recorded 2026-09-22
ImpalaLearner agents/runners/learners/impala_learner.py unverified MIT (permissive) · b3a1158468bd53aa · report
seolhokim/SimpleDistributedRL community (archive-listed) 2 samples, 0 ran · 2 unverified
get_vtrace_and_advantages distributed_reinforcement_learning/algorithms/impala.py unverified MIT (permissive) · b1527fc2ce13e346 · report
train in 522 papers distributed_reinforcement_learning/algorithms/impala.py unverified MIT (permissive) · 01b62761de35ed90 · report

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another.

11 assemblies on this paper: 11 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).

11 sibling assemblies for IMPALA (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for IMPALA (graph method: IMPALA). 2 bricks; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 11 siblings

  1. Impala entry · level 0 · unverified · cd20ae279ac1aa9f
  2. VTrace loss · level 1 · ran · a7ffabde1fe6a3c9

harvested Assembly for IMPALA (graph method: IMPALA). 3 bricks; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. from_logits entry · level 0 · unverified · c0ca8087e51d7e1d
  2. from_importance_weights helper · level 0 · unverified · 86f7ee994e64dc88
  3. log_probs_from_logits_and_actions helper · level 1 · ran; the contract check could not run (RAISES) · a872fa719c5242ff

harvested Assembly for IMPALA (graph method: IMPALA). 14 bricks; individually on a fixture: 7 ran (of which 3 constructed an object rather than computing a result; 7 with no instrument failure: 1 honoured, 1 violated, 5 with no contract checked; 0 where Syntology's instrument failed) · 7 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. AC entry · level 0 · unverified · f34df61f30b311d5
  2. ActionStateProbs helper · level 0 · unverified · 103308756a178abe
  3. CategoricalActor helper · level 1 · ran · e05eecb1a5750a6c
  4. Default helper · level 0 · unverified · ceddd134b3dd01f7
  5. PlayerState helper · level 0 · unverified · cbb6758024239ec3
  6. V model · level 1 · ran · 7d75b7a3afdfddc6
  7. nn_crossover helper · level 0 · unverified · e1569fbaca09c124
  8. Categorical helper · level 1 · ran · 89ed626ce6f952a2
  9. ConfigLoader helper · level 1 · ran · 4d9828f13202e523
  10. layer_crossover helper · level 0 · unverified · a22618d583e622fe
  11. pairwise_cross_corr helper · level 1 · ran · honoured contract · ba92bb44edd2ada6
  12. safe_crossover helper · level 1 · ran · violated contract · d9e1fb4c37c2e728
  13. Distribution helper · level 1 · ran · 6490dcf1e1570c31
  14. corr_matrix helper · level 0 · unverified · 9f5db323eec0eeba

harvested Assembly for IMPALA (graph method: IMPALA). 3 bricks; individually on a fixture: 3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. from_logits entry · level 1 · ran; the contract check could not run (RAISES) · 715b439b8761ea76
  2. action_log_probs helper · level 1 · ran · our draft was wrong · 815b225c5ab9f314
  3. from_importance_weights helper · level 1 · ran · our draft was wrong · 123172de8a1bc477

harvested Assembly for IMPALA (graph method: IMPALA). 1 brick; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 11 siblings

  1. GALA_A2C entry · level 1 · ran · aad00ea22af3c855

harvested Assembly for IMPALA (graph method: IMPALA). 1 brick; individually on a fixture: 0 ran · 1 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. calc_loss_and_backprop entry · level 0 · unverified · cb89e31851cc2931

harvested Assembly for IMPALA (graph method: IMPALA). 2 bricks; individually on a fixture: 1 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 1 unverified; the one sample that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 11 siblings

  1. ImpalaLearner entry · level 0 · unverified · b3a1158468bd53aa
  2. Learner helper · level 1 · ran · 3fdd0c84f214c2bb

harvested Assembly for IMPALA (graph method: IMPALA). 3 bricks; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. from_logits entry · level 0 · unverified · e26d17c678348c08
  2. action_log_probs helper · level 0 · unverified · bc2ac55256ed974b
  3. from_importance_weights helper · level 1 · ran · our draft was wrong · 123172de8a1bc477

harvested Assembly for IMPALA (graph method: IMPALA). 2 bricks; individually on a fixture: 0 ran · 2 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. train entry · level 0 · unverified · 01b62761de35ed90
  2. get_vtrace_and_advantages loss · level 0 · unverified · b1527fc2ce13e346

harvested Assembly for IMPALA (graph method: IMPALA). 1 brick; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. vtrace entry · level 1 · ran; the contract check could not run (RAISES) · e565e1e08272948c

harvested Assembly for IMPALA (graph method: IMPALA). 3 bricks; individually on a fixture: 1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 2 unverified. The assembly has not been run end-to-end.

one of 11 siblings

  1. from_logits entry · level 0 · unverified · 2b5d90526b4ad312
  2. from_importance_weights helper · level 0 · unverified · e11ee068514828e6
  3. log_probs_from_logits_and_actions helper · level 1 · ran; the contract check could not run (RAISES) · a75de4dc4c61d7dc

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="1802.01561") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Atari GamesReinforcement LearningReinforcement Learning (RL)reinforcement-learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Atari Games Atari 2600 Alien IMPALA (deep) Score 15962.10 #10 of 49 Archive leaderboard report
Atari Games Atari 2600 Amidar IMPALA (deep) Score 1554.79 #17 of 48 Archive leaderboard report
Atari Games Atari 2600 Assault IMPALA (deep) Score 19148.47 #11 of 45 Archive leaderboard report
Atari Games Atari 2600 Asterix IMPALA (deep) Score 300732.00 #16 of 49 Archive leaderboard report
Atari Games Atari 2600 Asteroids IMPALA (deep) Score 108590.05 #9 of 43 Archive leaderboard report
Atari Games Atari 2600 Atlantis IMPALA (deep) Score 849967.50 #21 of 42 Archive leaderboard report
Atari Games Atari 2600 Bank Heist IMPALA (deep) Score 1223.15 #16 of 45 Archive leaderboard report
Atari Games Atari 2600 Battle Zone IMPALA (deep) Score 20885.00 #36 of 47 Archive leaderboard report
Atari Games Atari 2600 Beam Rider IMPALA (deep) Score 32463.47 #12 of 49 Archive leaderboard report
Atari Games Atari 2600 Berzerk IMPALA (deep) Score 1852.70 #18 of 39 Archive leaderboard report
Atari Games Atari 2600 Bowling IMPALA (deep) Score 59.92 #25 of 44 Archive leaderboard report
Atari Games Atari 2600 Boxing IMPALA (deep) Score 99.96 #10 of 45 Archive leaderboard report
Atari Games Atari 2600 Breakout IMPALA (deep) Score 787.34 #10 of 58 Archive leaderboard report
Atari Games Atari 2600 Centipede IMPALA (deep) Score 11049.75 #18 of 45 Archive leaderboard report
Atari Games Atari 2600 Chopper Command IMPALA (deep) Score 28255.00 #12 of 45 Archive leaderboard report
Atari Games Atari 2600 Crazy Climber IMPALA (deep) Score 136950.00 #22 of 49 Archive leaderboard report
Atari Games Atari 2600 Defender IMPALA (deep) Score 185203.00 #11 of 21 Archive leaderboard report
Atari Games Atari 2600 Demon Attack IMPALA (deep) Score 132826.98 #10 of 46 Archive leaderboard report
Atari Games Atari 2600 Double Dunk IMPALA (deep) Score -0.33 #38 of 43 Archive leaderboard report
Atari Games Atari 2600 Enduro IMPALA (deep) Score 0.00 #45 of 48 Archive leaderboard report
Atari Games Atari 2600 Fishing Derby IMPALA (deep) Score 44.85 #15 of 44 Archive leaderboard report
Atari Games Atari 2600 Freeway IMPALA (deep) Score 0.00 #59 of 59 Archive leaderboard report
Atari Games Atari 2600 Frostbite IMPALA (deep) Score 317.75 #44 of 53 Archive leaderboard report
Atari Games Atari 2600 Gopher IMPALA (deep) Score 66782.30 #15 of 43 Archive leaderboard report
Atari Games Atari 2600 Gravitar IMPALA (deep) Score 359.50 #40 of 53 Archive leaderboard report
Atari Games Atari 2600 HERO IMPALA (deep) Score 33730.55 #9 of 44 Archive leaderboard report
Atari Games Atari 2600 Ice Hockey IMPALA (deep) Score 3.48 #14 of 43 Archive leaderboard report
Atari Games Atari 2600 James Bond IMPALA (deep) Score 601.50 #31 of 45 Archive leaderboard report
Atari Games Atari 2600 Kangaroo IMPALA (deep) Score 1632.00 #35 of 47 Archive leaderboard report
Atari Games Atari 2600 Krull IMPALA (deep) Score 8147.40 #28 of 45 Archive leaderboard report
Atari Games Atari 2600 Kung-Fu Master IMPALA (deep) Score 43375.50 #19 of 43 Archive leaderboard report
Atari Games Atari 2600 Montezuma's Revenge IMPALA (deep) Score 0.00 #44 of 50 Archive leaderboard report
Atari Games Atari 2600 Ms. Pacman IMPALA (deep) Score 7342.32 #12 of 47 Archive leaderboard report
Atari Games Atari 2600 Name This Game IMPALA (deep) Score 21537.20 #11 of 43 Archive leaderboard report
Atari Games Atari 2600 Phoenix IMPALA (deep) Score 210996.45 #10 of 21 Archive leaderboard report
Atari Games Atari 2600 Pitfall! IMPALA (deep) Score -1.66 #22 of 23 Archive leaderboard report
Atari Games Atari 2600 Pong IMPALA (deep) Score 20.98 #15 of 52 Archive leaderboard report
Atari Games Atari 2600 Private Eye IMPALA (deep) Score 98.50 #48 of 52 Archive leaderboard report
Atari Games Atari 2600 Q*Bert IMPALA (deep) Score 351200.12 #4 of 57 Archive leaderboard report
Atari Games Atari 2600 River Raid IMPALA (deep) Score 29608.05 #6 of 41 Archive leaderboard report
Atari Games Atari 2600 Road Runner IMPALA (deep) Score 57121.00 #19 of 44 Archive leaderboard report
Atari Games Atari 2600 Robotank IMPALA (deep) Score 12.96 #36 of 42 Archive leaderboard report
Atari Games Atari 2600 Seaquest IMPALA (deep) Score 1753.20 #41 of 57 Archive leaderboard report
Atari Games Atari 2600 Skiing IMPALA (deep) Score -10180.38 #6 of 23 Archive leaderboard report
Atari Games Atari 2600 Solaris IMPALA (deep) Score 2365.00 #19 of 23 Archive leaderboard report
Atari Games Atari 2600 Space Invaders IMPALA (deep) Score 43595.78 #9 of 55 Archive leaderboard report
Atari Games Atari 2600 Star Gunner IMPALA (deep) Score 200625.00 #8 of 43 Archive leaderboard report
Atari Games Atari 2600 Surround IMPALA (deep) Score 7.56 #8 of 15 Archive leaderboard report
Atari Games Atari 2600 Tennis IMPALA (deep) Score 0.55 #19 of 43 Archive leaderboard report
Atari Games Atari 2600 Time Pilot IMPALA (deep) Score 48481.50 #10 of 44 Archive leaderboard report
Atari Games Atari 2600 Tutankham IMPALA (deep) Score 292.11 #11 of 44 Archive leaderboard report
Atari Games Atari 2600 Up and Down IMPALA (deep) Score 332546.75 #11 of 44 Archive leaderboard report
Atari Games Atari 2600 Venture IMPALA (deep) Score 0.00 #52 of 55 Archive leaderboard report
Atari Games Atari 2600 Video Pinball IMPALA (deep) Score 572898.27 #13 of 42 Archive leaderboard report
Atari Games Atari 2600 Wizard of Wor IMPALA (deep) Score 9157.50 #22 of 41 Archive leaderboard report
Atari Games Atari 2600 Yars Revenge IMPALA (deep) Score 84231.14 #11 of 17 Archive leaderboard report
Atari Games Atari 2600 Zaxxon IMPALA (deep) Score 32935.50 #9 of 41 Archive leaderboard report
Atari Games Atari games IMPALA, deep Mean Human Normalized Score 957.34% #9 of 12 Archive leaderboard report
Atari Games Atari-57 IMPALA, deep Human World Record Breakthrough 3 #8 of 11 Archive leaderboard report
Atari Games Atari-57 IMPALA, deep Mean Human Normalized Score 957.34% #8 of 11 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Introduced by this paper: IMPALA, V-trace

ConvolutionEntropy RegularizationExperience ReplayGradient ClippingIMPALALSTMMax PoolingRMSPropReLUResidual ConnectionSigmoid ActivationTanh ActivationV-trace

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections