Papers › Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for...

Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning

17 Dec 2024arXiv:2412.12953archive 2025-07-28

Moritz Reuss, Jyothish Pari, Pulkit Agrawal, Rudolf Lioutikov

Diffusion Policies have become widely used in Imitation Learning, offering several appealing properties, such as generating multimodal and discontinuous behavior. As models are becoming larger to capture more complex capabilities, their computational demands increase, as shown by recent scaling laws. Therefore, continuing with the current architectures will present a computational roadblock. To address this gap, we propose Mixture-of-Denoising Experts (MoDE) as a novel policy for Imitation Learning. MoDE surpasses current state-of-the-art Transformer-based Diffusion Policies while enabling parameter-efficient scaling through sparse experts and noise-conditioned routing, reducing both active parameters by 40% and inference costs by 90% via expert caching. Our architecture combines this efficient scaling with noise-conditioned self-attention mechanism, enabling more effective denoising across different noise levels. MoDE achieves state-of-the-art performance on 134 tasks in four established imitation learning benchmarks (CALVIN and LIBERO). Notably, by pretraining MoDE on diverse robotics data, we achieve 4.01 on CALVIN ABC and 0.95 on LIBERO-90. It surpasses both CNN-based and Transformer Diffusion Policies by an average of 57% across 4 benchmarks, while using 90% fewer FLOPs and fewer active parameters compared to default Diffusion Transformer architectures. Furthermore, we conduct comprehensive ablations on MoDE's components, providing insights for designing efficient and scalable Transformer architectures for Diffusion Policies. Code and demonstrations are available at https://mbreuss.github.io/MoDE_Diffusion_Policy/.

PaperPDFCodeCode Syntology ran

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 11 code samples harvested from 1 repository linked to this paper: 10 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 2 honoured, 0 violated, 5 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 2 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 11 samples from 1 repository, 10 ran (7 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. 1 of the 10 samples that ran has no recorded date.

intuitive-robots/MoDE_Diffusion_Policy mentioned on GitHubMITSyntology: 11 samples, 10 ran (3 instrument-failed) · 1 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

11 samples harvested: 10 ran (of which 0 constructed an object rather than computing a result; 7 with no instrument failure: 2 honoured, 0 violated, 5 with no contract checked; 3 where Syntology's instrument failed) · 1 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.

2ran · honoured contract
1ran · our draft was wrong
2ran; the contract check could not run
5ran
1unverified

Licence: 3 of the 11 samples are pointer only, meaning Syntology does not serve that copy's text. For 3 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from intuitive-robots/MoDE_Diffusion_Policy. “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. 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.

intuitive-robots/MoDE_Diffusion_Policy community (archive-listed) 11 samples, 10 ran (3 instrument-failed) · 1 unverified
append_dims in 30 papers
census: 3 implementations from 6 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
mode/models/edm_diffusion/utils.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted MIT (permissive) here; the stored copy is pointer only · 30befb7e4327e615 · report fixture: signature x:Tensor/2/float, target_dims:int/0/int · returned Tensor · level recorded, date not recorded
append_zero in 5 papers mode/models/edm_diffusion/gc_sampling.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted MIT (permissive) here; the stored copy is pointer only · 43f2167b019a5fb7 · report fixture: signature action:Tensor/1/float · returned Tensor · level recorded 2026-09-22
get_sigmas_exponential mode/models/edm_diffusion/gc_sampling.py ran · honoured contract MIT (permissive) · 205e97641ece09f8 · report fixture: signature n:int/0/int, sigma_min:float/0/float, sigma_max:float/0/float, device:str/0/str · returned Tensor · level recorded 2026-09-24
get_sigmas_karras in 8 papers mode/models/edm_diffusion/gc_sampling.py ran · honoured contract MIT (permissive) · a3ae2e8a897b4d99 · report fixture: signature n:int/0/int, sigma_min:float/0/float, sigma_max:float/0/float, rho:float/0/float, device:str/0/str · returned Tensor · level recorded 2026-09-24
load_npz in 17 papers mode/datasets/disk_dataset.py ran MIT (permissive) · 68787b0149b5e4bf · report fixture: signature filename:PosixPath/0/? · returned NpzFile · level recorded 2026-09-24
load_pkl in 46 papers mode/datasets/disk_dataset.py ran MIT (permissive) · 45af63efa9933397 · report fixture: signature filename:PosixPath/0/? · returned dict · level recorded 2026-09-24
rand_log_normal in 7 papers mode/models/edm_diffusion/utils.py ran · our draft was wrong MIT (permissive) here; the stored copy is pointer only · ec8364ee198f15b3 · report fixture: signature shape:tuple/1/int, loc:float/0/float, scale:float/0/float, device:str/0/str, dtype:dtype/0/? · returned Tensor · level recorded 2026-09-22
return_time_sigma_embedding_model mode/models/edm_diffusion/utils.py ran MIT (permissive) · 6f1fce523ddeea43 · report fixture: signature embedding_type:str/0/str, time_embed_dim:int/0/int, device:str/0/str · returned GaussianFourierEmbedding · level recorded 2026-09-24
setup_callbacks in 6 papers mode/training_calvin.py ran MIT (permissive) · 0e68bc534d7f6740 · report fixture: signature callbacks_cfg:DictConfig/0/? · returned list · level recorded 2026-09-25
split_trajectories mode/datasets/libero_data_module.py ran MIT (permissive) · 38618aea799ddc0e · report fixture: signature num_trajectories:int/0/int, train_ratio:float/0/float · returned tuple · level recorded 2026-09-25
setup_logger in 122 papers mode/training_calvin.py unverified MIT (permissive) · 765bd1c53cf958eb · report

Tasks

DenoisingImitation LearningRobot ManipulationZero-shot Generalization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
LIBERO-10 MoDE Average Success Rate 0.94 #1 of 1 Archive leaderboard report
LIBERO-90 MoDE Average Success Rate 0.95 #1 of 1 Archive leaderboard report
Robot Manipulation CALVIN MoDE avg. sequence length (D to D) 4.01 #7 of 19 Archive leaderboard report
Zero-shot Generalization CALVIN MoDE Avg. sequence length 4.01 #2 of 5 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

ABCAbsolute Position EncodingsAdamAttentionBPEDense ConnectionsDiffusionDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

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