Papers › Bridging MARL to SARL: An Order-Independent Multi-Agent Transformer via Latent Consensus

Bridging MARL to SARL: An Order-Independent Multi-Agent Transformer via Latent Consensus

15 Apr 2026arXiv:2604.13472added by Syntology

Zijian Zhao, Jing Gao, Sen Li

Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).

Cooperative multi-agent reinforcement learning (MARL) is widely used to address large joint observation and action spaces by decomposing a centralized control problem into multiple interacting agents. However, such decomposition often introduces additional challenges, including non-stationarity, unstable training, weak coordination, and limited theoretical guarantees. In this paper, we propose the Consensus Multi-Agent Transformer (CMAT), a centralized framework that bridges cooperative MARL to a hierarchical single-agent reinforcement learning (SARL) formulation. CMAT treats all agents as a unified entity and employs a Transformer encoder to process the large joint observation space. To handle the extensive joint action space, we introduce a hierarchical decision-making mechanism in which a Transformer decoder autoregressively generates a high-level consensus vector, simulating the process by which agents reach agreement on their strategies in latent space. Conditioned on this consensus, all agents generate their actions simultaneously, enabling order-independent joint decision making and avoiding the sensitivity to action-generation order in conventional Multi-Agent Transformers (MAT). This factorization allows the joint policy to be optimized using single-agent PPO while preserving expressive coordination through the latent consensus. To evaluate the proposed method, we conduct experiments on benchmark tasks from StarCraft II, Multi-Agent MuJoCo, and Google Research Football. The results show that CMAT achieves superior performance over recent centralized solutions, sequential MARL methods, and conventional MARL baselines. The code for this paper is available at:https://github.com/RS2002/CMAT .

PaperPDFCode 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 12 code samples harvested from 1 repository linked to this paper: 9 ran (of which 2 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 5 where Syntology's instrument failed) · 3 unverified. Where Syntology's instrument failed: 5 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: found in paper text by Syntology: 12 samples from 1 repository, 9 ran (4 with no instrument failure; 5 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 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

PKU-MARL/Multi-Agent-Transformer found in paper text by SyntologySyntology: 12 samples, 9 ran (5 instrument-failed) · 3 unverified report

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

12 samples harvested: 9 ran (of which 2 constructed an object rather than computing a result; 4 with no instrument failure: 0 honoured, 0 violated, 4 with no contract checked; 5 where Syntology's instrument failed) · 3 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.

5ran · our draft was wrong
4ran
3unverified

Licence: 12 of the 12 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 PKU-MARL/Multi-Agent-Transformer. “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 1 sample 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.

PKU-MARL/Multi-Agent-Transformer found in paper text by Syntology 12 samples, 9 ran (5 instrument-failed) · 3 unverified
DecodeBlock mat/algorithms/mat/algorithm/ma_transformer.py ran no licence file found · pointer only · 362c1be3708b32ec · report fixture: signature x:Tensor/3/float, rep_enc:Tensor/3/float · returned Tensor · level recorded 2026-09-24
EncodeBlock mat/algorithms/mat/algorithm/ma_transformer.py ran · metamorphic tier: deterministic no licence file found · pointer only · 38a9cb3379b34d27 · report fixture: returned EncodeBlock · level recorded 2026-09-22
Encoder in 278 papers mat/algorithms/mat/algorithm/ma_transformer.py ran no licence file found · pointer only · 9f99db24b15d292c · report fixture: signature state:Tensor/3/float, obs:Tensor/3/float · returned tuple · level recorded 2026-09-24
SelfAttention in 66 papers mat/algorithms/mat/algorithm/ma_transformer.py ran · metamorphic tier: deterministic no licence file found · pointer only · 508634b69e88baa8 · report fixture: returned SelfAttention · level recorded 2026-09-22
continuous_autoregreesive_act mat/algorithms/mat/algorithm/ma_transformer.py ran · our draft was wrong no licence file found · pointer only · 59cb0ba2cc3b36c6 · report fixture: signature decoder:callable/0/?, obs_rep:Tensor/2/float, obs:Tensor/2/float, batch_size:int/0/int, n_agent:int/0/int, action_dim:int/0/int, tpdv:dict/0/?, deterministic:bool/0/bool · returned tuple · level recorded 2026-09-22
continuous_parallel_act mat/algorithms/mat/algorithm/ma_transformer.py ran · our draft was wrong no licence file found · pointer only · 80d5ea44b6503397 · report fixture: signature decoder:callable/0/?, obs_rep:Tensor/3/float, obs:Tensor/3/float, action:Tensor/3/float, batch_size:int/0/int, n_agent:int/0/int, action_dim:int/0/int, tpdv:dict/0/? · returned tuple · level recorded 2026-09-22
discrete_parallel_act mat/algorithms/mat/algorithm/ma_transformer.py ran · our draft was wrong no licence file found · pointer only · cdc64fb18b542a6b · report fixture: signature decoder:callable/0/?, obs_rep:Tensor/3/float, obs:Tensor/3/float, action:Tensor/3/int, batch_size:int/0/int, n_agent:int/0/int, action_dim:int/0/int, tpdv:dict/0/?, available_actions:Tensor/3/float · returned tuple · level recorded 2026-09-22
init in 55 papers mat/algorithms/mat/algorithm/ma_transformer.py ran · our draft was wrong no licence file found · pointer only · c00885ffa5fe61e7 · report fixture: signature module:callable/0/?, weight_init:callable/0/?, bias_init:callable/0/?, gain:int/0/int · returned SimpleModule · level recorded 2026-09-22
init_ in 12 papers mat/algorithms/mat/algorithm/ma_transformer.py ran · our draft was wrong no licence file found · pointer only · 748385c8184d4945 · report fixture: signature m:callable/0/?, gain:float/0/float, activate:bool/0/bool · returned Linear · level recorded 2026-09-22
Decoder in 210 papers mat/algorithms/mat/algorithm/ma_transformer.py unverified no licence file found · pointer only · da968f0f2c007dff · report
MultiAgentTransformer mat/algorithms/mat/algorithm/ma_transformer.py unverified no licence file found · pointer only · 8aa233770f15e845 · report
discrete_autoregreesive_act mat/algorithms/mat/algorithm/ma_transformer.py unverified no licence file found · pointer only · 55cd239026a74d42 · 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.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for Consensus Multi-Agent Transformer (CMAT) (graph method: Consensus Multi-Agent Transformer (CMAT)). 13 bricks; individually on a fixture: 10 ran (of which 2 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 0 violated, 4 with no contract checked; 5 where Syntology's instrument failed) · 3 unverified. The assembly has not been run end-to-end.

  1. MultiAgentTransformer entry · level 0 · unverified · 8aa233770f15e845
  2. Decoder model · level 0 · unverified · da968f0f2c007dff
  3. Encoder model · level 1 · ran · 9f99db24b15d292c
  4. check helper · level 1 · ran · honoured contract · 4829105008c71d83
  5. continuous_autoregreesive_act helper · level 1 · ran · our draft was wrong · 59cb0ba2cc3b36c6
  6. continuous_parallel_act helper · level 1 · ran · our draft was wrong · 80d5ea44b6503397
  7. discrete_autoregreesive_act helper · level 0 · unverified · 55cd239026a74d42
  8. discrete_parallel_act helper · level 1 · ran · our draft was wrong · cdc64fb18b542a6b
  9. DecodeBlock model · level 1 · ran · 362c1be3708b32ec
  10. init_ helper · level 1 · ran · our draft was wrong · 748385c8184d4945
  11. EncodeBlock model · level 1 · ran · 38a9cb3379b34d27
  12. SelfAttention model · level 1 · ran · 508634b69e88baa8
  13. init helper · level 1 · ran · our draft was wrong · c00885ffa5fe61e7

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="2604.13472") returns these assemblies with each brick's role and wiring (how to connect).

Results from the paper

The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2604.13472, was issued in April 2026, after that date, so the archive has no leaderboard rows for it.

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