Papers › ResMLP: Feedforward networks for image classification with data-efficient training

ResMLP: Feedforward networks for image classification with data-efficient training

7 May 2021NeurIPS 2021 12arXiv:2105.03404archive 2025-07-28

Hugo Touvron, Piotr Bojanowski, Mathilde Caron, Matthieu Cord, Alaaeldin El-Nouby, Edouard Grave, Gautier Izacard, Armand Joulin, Gabriel Synnaeve, Jakob Verbeek, Hervé Jégou

We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. It is a simple residual network that alternates (i) a linear layer in which image patches interact, independently and identically across channels, and (ii) a two-layer feed-forward network in which channels interact independently per patch. When trained with a modern training strategy using heavy data-augmentation and optionally distillation, it attains surprisingly good accuracy/complexity trade-offs on ImageNet. We also train ResMLP models in a self-supervised setup, to further remove priors from employing a labelled dataset. Finally, by adapting our model to machine translation we achieve surprisingly good results. We share pre-trained models and our code based on the Timm library.

PaperPDFConference PDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2105.03404")

Code

Syntology Ran 2 of 7 code samples harvested from 4 repositories linked to this paper; 5 have no recorded run. Of those that ran: 1 ran · honoured contract; 1 ran · violated contract.

By repository: community (archive-listed): 7 samples from 4 repositories, 2 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

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

facebookresearch/deit officialpytorch report
rwightman/pytorch-image-models mentioned in papermentioned on GitHubpytorch report
IMvision12/keras-vision-models mentioned on GitHubpytorch report
Mayurji/Image-Classification-PyTorch mentioned on GitHubpytorch report
jaketae/res-mlp mentioned on GitHubpytorch report
lalithjets/surgical_vqa mentioned on GitHubpytorch report
leaderj1001/Bag-of-MLP mentioned on GitHubpytorch report
liuruiyang98/Jittor-MLP mentioned on GitHubjax report
lucidrains/res-mlp-pytorch mentioned on GitHubpytorch report
rishikksh20/ResMLP-pytorch mentioned on GitHubpytorchMIT 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

7 samples harvested; 2 ran; 1 honoured the contract we drafted; 5 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · honoured contract
1ran · violated contract
5unverified

Licence: 0 of the 7 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 4 repositories linked to this paper, official or community; each sample names its own and says which. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

check_sizes jaketae/res-mlp/res_mlp/core.py community (archive-listed) ran · honoured contract fingerprinted MIT (permissive) · 27cbea18b00696e9 · report
pair lucidrains/res-mlp-pytorch/res_mlp_pytorch/res_mlp_pytorch.py community (archive-listed) ran · violated contract fingerprinted MIT (permissive) · e6f9d3b5ad66ee20 · report
PatchEmbed yeyinthtoon/tf2-resmlp/resmlp/layers.py community (archive-listed) unverified MIT (permissive) · c5a2d135cecde44d · report
adjust_block_compatibility megvii-research/basecls/basecls/layers/wrapper.py community (archive-listed) unverified Apache-2.0 (permissive) · 6a660e5991846119 · report
layers_scale_mlp_blocks yeyinthtoon/tf2-resmlp/resmlp/resmlp.py community (archive-listed) unverified MIT (permissive) · 3099f730f63abb47 · report
mlp_block yeyinthtoon/tf2-resmlp/resmlp/resmlp.py community (archive-listed) unverified MIT (permissive) · 6c615e583ce7944d · report
resmlp yeyinthtoon/tf2-resmlp/resmlp/resmlp.py community (archive-listed) unverified MIT (permissive) · 87c6692221a5e172 · report

Tasks

Data AugmentationFine-Grained Image ClassificationGeneral ClassificationImage ClassificationMachine TranslationSelf-Supervised Image ClassificationTranslationimage-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Fine-Grained Image Classification Oxford 102 Flowers ResMLP-24 Accuracy 97.9% #17 of 25 Archive leaderboard report
Fine-Grained Image Classification Oxford 102 Flowers ResMLP-12 Accuracy 97.4% #19 of 25 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars ResMLP-24 Accuracy 89.5% #79 of 83 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars ResMLP-12 Accuracy 84.6% #81 of 83 Archive leaderboard report
Image Classification CIFAR-100 ResMLP-24 Percentage correct 89.5 #29 of 211 Archive leaderboard report
Image Classification CIFAR-100 ResMLP-12 Percentage correct 87.0 #47 of 211 Archive leaderboard report
Image Classification Certificate Verification ResMLP-24 Percentage correct 98.7 #1 of 2 Archive leaderboard report
Image Classification Certificate Verification ResMLP-24 Top-1 Accuracy 98.7 #1 of 2 Archive leaderboard report
Image Classification Certificate Verification ResMLP-12 Percentage correct 98.1 #2 of 2 Archive leaderboard report
Image Classification Certificate Verification ResMLP-12 Top-1 Accuracy 98.1 #2 of 2 Archive leaderboard report
Image Classification Flowers-102 ResMLP24 Accuracy 97.9 #32 of 52 Archive leaderboard report
Image Classification Flowers-102 ResMLP12 Accuracy 97.4 #38 of 52 Archive leaderboard report
Image Classification ImageNet ResMLP-B24/8 Number of params 116M #418 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-B24/8 Top 1 Accuracy 83.6% #418 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-S24 GFLOPs 6 #683 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-S24 Number of params 30M #683 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-S24 Top 1 Accuracy 80.8% #683 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-36 Number of params 45M #746 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-36 Top 1 Accuracy 79.7% #746 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-24 Top 1 Accuracy 79.4% #757 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-12 (distilled, class-MLP) GFLOPs 3 #822 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-12 (distilled, class-MLP) Number of params 17.7M #822 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-12 (distilled, class-MLP) Top 1 Accuracy 78.6% #822 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-S12 Number of params 15.4M #864 of 1060 Archive leaderboard report
Image Classification ImageNet ResMLP-S12 Top 1 Accuracy 77.8% #864 of 1060 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-36 Accuracy 85.6% #41 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-36 Params 45M #41 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-24 Accuracy 85.3% #42 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-24 Params 30M #42 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-12 Accuracy 84.6% #44 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-12 Params 15M #44 of 57 Archive leaderboard report
Image Classification ImageNet ReaL ResMLP-B24/8 (22k) Top 1 Accuracy 84.4% #57 of 57 Archive leaderboard report
Image Classification ImageNet V2 ResMLP-B24/8 22k Top 1 Accuracy 74.2 #18 of 33 Archive leaderboard report
Image Classification ImageNet V2 ResMLP-B24/8 Top 1 Accuracy 73.4 #20 of 33 Archive leaderboard report
Image Classification ImageNet V2 ResMLP-S24/16 Top 1 Accuracy 69.8 #25 of 33 Archive leaderboard report
Image Classification ImageNet V2 ResMLP-S12/16 Top 1 Accuracy 66.0 #31 of 33 Archive leaderboard report
Image Classification Stanford Cars ResMLP-24 Accuracy 89.5 #16 of 24 Archive leaderboard report
Image Classification Stanford Cars ResMLP-12 Accuracy 84.6 #22 of 24 Archive leaderboard report
Image Classification iNaturalist 2018 ResMLP-24 Top-1 Accuracy 64.3 #51 of 60 Archive leaderboard report
Image Classification iNaturalist 2018 ResMLP-12 Top-1 Accuracy 60.2 #55 of 60 Archive leaderboard report
Image Classification iNaturalist 2019 ResMLP-24 Top-1 Accuracy 72.5 #17 of 22 Archive leaderboard report
Image Classification iNaturalist 2019 ResMLP-12 Top-1 Accuracy 71.0 #19 of 22 Archive leaderboard report
Machine Translation WMT2014 English-French ResMLP-12 BLEU score 40.6 #30 of 57 Archive leaderboard report
Machine Translation WMT2014 English-French ResMLP-6 BLEU score 40.3 #34 of 57 Archive leaderboard report
Machine Translation WMT2014 English-German ResMLP-12 BLEU score 26.8 #55 of 91 Archive leaderboard report
Machine Translation WMT2014 English-German ResMLP-6 BLEU score 26.4 #59 of 91 Archive leaderboard report
Self-Supervised Image Classification ImageNet DINO (ResMLP-24) Number of Params 30M #93 of 144 Archive leaderboard report
Self-Supervised Image Classification ImageNet DINO (ResMLP-24) Top 1 Accuracy 72.8% #93 of 144 Archive leaderboard report
Self-Supervised Image Classification ImageNet DINO (ResMLP-12) Number of Params 15M #109 of 144 Archive leaderboard report
Self-Supervised Image Classification ImageNet DINO (ResMLP-12) Top 1 Accuracy 67.5% #109 of 144 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: Class-MLP, ResMLP

Affine OperatorClass-MLPDense ConnectionsFeedforward NetworkLayerScaleLinear LayerResMLPResidual Connection

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