Papers › Beyond The Rainbow: High Performance Deep Reinforcement Learning on a Desktop PC

Beyond The Rainbow: High Performance Deep Reinforcement Learning on a Desktop PC

6 Nov 2024arXiv:2411.03820archive 2025-07-28

Tyler Clark, Mark Towers, Christine Evers, Jonathon Hare

Rainbow Deep Q-Network (DQN) demonstrated combining multiple independent enhancements could significantly boost a reinforcement learning (RL) agent's performance. In this paper, we present "Beyond The Rainbow" (BTR), a novel algorithm that integrates six improvements from across the RL literature to Rainbow DQN, establishing a new state-of-the-art for RL using a desktop PC, with a human-normalized interquartile mean (IQM) of 7.4 on Atari-60. Beyond Atari, we demonstrate BTR's capability to handle complex 3D games, successfully training agents to play Super Mario Galaxy, Mario Kart, and Mortal Kombat with minimal algorithmic changes. Designing BTR with computational efficiency in mind, agents can be trained using a high-end desktop PC on 200 million Atari frames within 12 hours. Additionally, we conduct detailed ablation studies of each component, analyzing the performance and impact using numerous measures. Code is available at https://github.com/VIPTankz/BTR.

PaperPDFCodeCode 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

Syntology 25 code samples harvested from 3 repositories linked to this paper: 16 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; 4 where Syntology's instrument failed) · 9 unverified. Where Syntology's instrument failed: 4 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 4 samples from 1 repository, 3 ran (2 with no instrument failure; 1 where Syntology's instrument failed); community (archive-listed): 21 samples from 2 repositories, 13 ran (10 with no instrument failure; 3 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.

Levels written between 2026-09-22 and 2026-09-25 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

viptankz/btr officialmentioned in paperpytorchMITSyntology: 4 samples, 3 ran (1 instrument-failed) · 1 unverified report
rickyegl/BeyondTheRainbowICLR pytorchSyntology: 18 samples, 13 ran (3 instrument-failed) · 5 unverified report
rickyegl/BeyondTheRainbowM pytorchSyntology: 3 samples, 0 ran · 3 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.

Code Syntology ran Syntology

25 samples harvested: 16 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; 4 where Syntology's instrument failed) · 9 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

4ran · our draft was wrong
12ran
9unverified

Licence: 21 of the 25 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 3 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”. For 4 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

viptankz/btr official repository 4 samples, 3 ran (1 instrument-failed) · 1 unverified
choose_eval_action Agent.py ran MIT (permissive) · 5702ceae0fbebc7b · report fixture: signature observation:Tensor/2/float, eval_net:callable/0/?, n_actions:int/0/int, device:str/0/str, rng:float/0/float · returned Tensor · level recorded 2026-09-25
moving_average in 46 papers
census: 9 implementations from 11 papers, 6 distinct outputs across 4 buckets, one shared input per bucket
Analytic.py ran · our draft was wrong fingerprinted MIT (permissive) · b47df527aff11259 · report fixture: signature data:ndarray/1/float, window_size:int/0/int · returned ndarray · level recorded 2026-09-25
randomise_action_batch Agent.py ran MIT (permissive) · 48242805f8b06f3a · report fixture: signature x:Tensor/2/int, probs:float/0/float, n_actions:int/0/int · returned Tensor · level recorded 2026-09-25
create_experience PER.py unverified MIT (permissive) · b7f6c4806f2bb9ff · report
rickyegl/BeyondTheRainbowICLR community (archive-listed) 18 samples, 13 ran (3 instrument-failed) · 5 unverified
Analytics Agent.py ran no licence file found · pointer only · ebfa2628b0000e65 · report fixture: returned Analytics · level recorded 2026-09-22
Dueling Agent.py ran fingerprinted no licence file found · pointer only · bacc2cddb664bea4 · report fixture: returned Dueling · level recorded 2026-09-22
EpsilonGreedy Agent.py ran no licence file found · pointer only · 17e8fcddd3188058 · report fixture: returned EpsilonGreedy · level recorded 2026-09-22
FactorizedNoisyLinear Agent.py ran no licence file found · pointer only · 4d142f883464d0f9 · report fixture: returned FactorizedNoisyLinear · level recorded 2026-09-22
ImpalaCNNBlock Agent.py ran fingerprinted no licence file found · pointer only · 888002b8a86e5a03 · report fixture: returned ImpalaCNNBlock · level recorded 2026-09-22
ImpalaCNNLargeC51 Agent.py ran no licence file found · pointer only · e9e1b699914b69e5 · report fixture: returned ImpalaCNNLargeC51 · level recorded 2026-09-22
ImpalaCNNResidual Agent.py ran no licence file found · pointer only · 41b0ef198d33a0be · report fixture: returned ImpalaCNNResidual · level recorded 2026-09-22
NoisyLinear in 9 papers Agent.py ran no licence file found · pointer only · 9e6f2536b7248907 · report fixture: returned NoisyLinear · level recorded 2026-09-22
PER Agent.py ran no licence file found · pointer only · 5cbd6045f5cc87d7 · report fixture: returned PER · level recorded 2026-09-22
SumTree Agent.py ran no licence file found · pointer only · 20c72900e5bac6d6 · report fixture: returned SumTree · level recorded 2026-09-22
calculate_huber_loss in 5 papers Agent.py ran · our draft was wrong no licence file found · pointer only · 7c1d5bc9f897b0cf · report fixture: signature td_errors:Tensor/3/float, k:float/0/float, taus:int/0/int · returned Tensor · level recorded 2026-09-22
distr_projection Agent.py ran · our draft was wrong no licence file found · pointer only · 8f1b5f746f353929 · report fixture: signature next_distr:Tensor/2/float, rewards:Tensor/1/float, dones:Tensor/1/bool, Vmin:float/0/float, Vmax:float/0/float, n_atoms:int/0/int, gamma:float/0/float · returned Tensor · level recorded 2026-09-22
randomise_action_batch Agent.py ran · our draft was wrong no licence file found · pointer only · 421450b8959b8fd6 · report fixture: signature x:Tensor/2/int, probs:float/0/float, n_actions:int/0/int · returned Tensor · level recorded 2026-09-22
Agent in 42 papers Agent.py unverified no licence file found · pointer only · 286da1c5eb5d8dde · report
ImpalaCNNLarge Agent.py unverified no licence file found · pointer only · 5b8f960bf199e9b6 · report
ImpalaCNNLargeIQN Agent.py unverified no licence file found · pointer only · 2fda2e0b0dc97a26 · report
NatureIQN Agent.py unverified no licence file found · pointer only · adf5a79486613798 · report
create_network in 7 papers Agent.py unverified no licence file found · pointer only · 0a3ccb7803118980 · report
rickyegl/BeyondTheRainbowM community (archive-listed) 3 samples, 0 ran · 3 unverified
Agent in 42 papers Agent.py unverified no licence file found · pointer only · 45091c96548faa63 · report
ImpalaCNNLargeIQN Agent.py unverified no licence file found · pointer only · 63f4b89e2f29bee0 · report
create_network in 7 papers Agent.py unverified no licence file found · pointer only · de954c232508133f · 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.

2 assemblies on this paper: 2 harvested, 0 generated, listed in 1 groups (siblings lifted in one batch form one group).

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

harvested Assembly for Beyond The Rainbow (BTR) (graph method: Beyond The Rainbow (BTR)). 18 bricks; individually on a fixture: 13 ran (of which 10 constructed an object rather than computing a result; 10 with no instrument failure: 0 honoured, 0 violated, 10 with no contract checked; 3 where Syntology's instrument failed) · 5 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. Agent entry · level 0 · unverified · 286da1c5eb5d8dde
  2. Analytics helper · level 1 · ran · ebfa2628b0000e65
  3. EpsilonGreedy sampler · level 1 · ran · 17e8fcddd3188058
  4. FactorizedNoisyLinear model · level 1 · ran · 4d142f883464d0f9
  5. PER data_transform · level 1 · ran · 5cbd6045f5cc87d7
  6. calculate_huber_loss loss · level 1 · ran · our draft was wrong · 7c1d5bc9f897b0cf
  7. create_network model · level 0 · unverified · 0a3ccb7803118980
  8. distr_projection update_rule · level 1 · ran · our draft was wrong · 8f1b5f746f353929
  9. randomise_action_batch sampler · level 1 · ran · our draft was wrong · 421450b8959b8fd6
  10. SumTree data_transform · level 1 · ran · 20c72900e5bac6d6
  11. ImpalaCNNLarge model · level 0 · unverified · 5b8f960bf199e9b6
  12. ImpalaCNNLargeC51 model · level 1 · ran · e9e1b699914b69e5
  13. ImpalaCNNLargeIQN model · level 0 · unverified · 2fda2e0b0dc97a26
  14. NatureIQN model · level 0 · unverified · adf5a79486613798
  15. Dueling model · level 1 · ran · bacc2cddb664bea4
  16. ImpalaCNNBlock model · level 1 · ran · 888002b8a86e5a03
  17. NoisyLinear model · level 1 · ran · 9e6f2536b7248907
  18. ImpalaCNNResidual model · level 1 · ran · 41b0ef198d33a0be

harvested Assembly for Beyond The Rainbow (BTR) (graph method: Beyond The Rainbow (BTR)). 14 bricks; individually on a fixture: 11 ran (of which 8 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 3 where Syntology's instrument failed) · 3 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. Agent entry · level 0 · unverified · 45091c96548faa63
  2. Analytics helper · level 1 · ran · ebfa2628b0000e65
  3. EpsilonGreedy helper · level 1 · ran · 17e8fcddd3188058
  4. FactorizedNoisyLinear model · level 1 · ran · 4d142f883464d0f9
  5. PER sampler · level 1 · ran · 5cbd6045f5cc87d7
  6. calculate_huber_loss loss · level 1 · ran · our draft was wrong · 7c1d5bc9f897b0cf
  7. create_network model · level 0 · unverified · de954c232508133f
  8. distr_projection update_rule · level 1 · ran · our draft was wrong · 8f1b5f746f353929
  9. randomise_action_batch helper · level 1 · ran · our draft was wrong · 421450b8959b8fd6
  10. SumTree helper · level 1 · ran · 20c72900e5bac6d6
  11. ImpalaCNNLargeIQN model · level 0 · unverified · 63f4b89e2f29bee0
  12. Dueling model · level 1 · ran · bacc2cddb664bea4
  13. ImpalaCNNBlock helper · level 1 · ran · 888002b8a86e5a03
  14. ImpalaCNNResidual helper · level 1 · ran · 41b0ef198d33a0be

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="2411.03820") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Computational EfficiencyDeep Reinforcement LearningReinforcement Learning (RL)

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

ConvolutionDQNDense ConnectionsDouble Q-learningDueling NetworkN-step ReturnsNoisy Linear LayerPrioritized Experience ReplayQ-LearningRainbow DQN

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