Papers › Scaling Vision with Sparse Mixture of Experts

Scaling Vision with Sparse Mixture of Experts

10 Jun 2021NeurIPS 2021 12arXiv:2106.05974archive 2025-07-28

Carlos Riquelme, Joan Puigcerver, Basil Mustafa, Maxim Neumann, Rodolphe Jenatton, André Susano Pinto, Daniel Keysers, Neil Houlsby

Sparsely-gated Mixture of Experts networks (MoEs) have demonstrated excellent scalability in Natural Language Processing. In Computer Vision, however, almost all performant networks are "dense", that is, every input is processed by every parameter. We present a Vision MoE (V-MoE), a sparse version of the Vision Transformer, that is scalable and competitive with the largest dense networks. When applied to image recognition, V-MoE matches the performance of state-of-the-art networks, while requiring as little as half of the compute at inference time. Further, we propose an extension to the routing algorithm that can prioritize subsets of each input across the entire batch, leading to adaptive per-image compute. This allows V-MoE to trade-off performance and compute smoothly at test-time. Finally, we demonstrate the potential of V-MoE to scale vision models, and train a 15B parameter model that attains 90.35% on ImageNet.

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="2106.05974")

Code

Syntology Ran 1 of 1 code samples harvested from 1 repository linked to this paper; 0 have no recorded run. Of those that ran: 1 ran · honoured contract.

By repository: official repository: 1 sample from 1 repository, 1 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

google-research/vmoe officialmentioned on GitHubjax 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

1 sample harvested; 1 ran; 1 honoured the contract we drafted; 0 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

Licence: 0 of the 1 sample 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 google-research/vmoe. “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.

get_mixup_concentration google-research/vmoe/vmoe/configs/vmoe_paper/common.py official repository ran · honoured contract Apache-2.0 (permissive) · b7c6719771038ec2 · report

Tasks

Few-Shot Image ClassificationImage ClassificationMixture-of-Experts

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Few-Shot Image Classification ImageNet - 1-shot ViT-MoE-15B (Every-2) Top 1 Accuracy 68.66 #1 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 1-shot V-MoE-H/14 (Every-2) Top 1 Accuracy 63.38 #3 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 1-shot V-MoE-H/14 (Last-5) Top 1 Accuracy 62.95 #4 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 1-shot V-MoE-L/16 (Every-2) Top 1 Accuracy 62.41 #5 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 1-shot VIT-H/14 Top 1 Accuracy 62.34 #6 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 10-shot ViT-MoE-15B (Every-2) Top 1 Accuracy 84.29 #2 of 7 Archive leaderboard report
Few-Shot Image Classification ImageNet - 10-shot V-MoE-H/14 (Every-2) Top 1 Accuracy 80.33 #5 of 7 Archive leaderboard report
Few-Shot Image Classification ImageNet - 10-shot V-MoE-H/14 (Last-5) Top 1 Accuracy 80.1 #6 of 7 Archive leaderboard report
Few-Shot Image Classification ImageNet - 10-shot VIT-H/14 Top 1 Accuracy 79.01 #7 of 7 Archive leaderboard report
Few-Shot Image Classification ImageNet - 5-shot ViT-MoE-15B (Every-2) Top 1 Accuracy 82.78 #1 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 5-shot V-MoE-H/14 (Every-2) Top 1 Accuracy 78.21 #5 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 5-shot V-MoE-H/14 (Last-5) Top 1 Accuracy 78.08 #6 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 5-shot V-MoE-L/16 (Every-2) Top 1 Accuracy 77.1 #7 of 8 Archive leaderboard report
Few-Shot Image Classification ImageNet - 5-shot VIT-H/14 Top 1 Accuracy 76.95 #8 of 8 Archive leaderboard report
Image Classification ImageNet V-MoE-H/14 (Every-2) Number of params 7200M #52 of 1060 Archive leaderboard report
Image Classification ImageNet V-MoE-H/14 (Every-2) Top 1 Accuracy 88.36% #52 of 1060 Archive leaderboard report
Image Classification ImageNet VIT-H/14 Number of params 656M #61 of 1060 Archive leaderboard report
Image Classification ImageNet VIT-H/14 Top 1 Accuracy 88.08% #61 of 1060 Archive leaderboard report
Image Classification ImageNet V-MoE-L/16 (Every-2) Number of params 3400M #88 of 1060 Archive leaderboard report
Image Classification ImageNet V-MoE-L/16 (Every-2) Top 1 Accuracy 87.41% #88 of 1060 Archive leaderboard report
Image Classification JFT-300M V-MoE-H/14 (Every-2) prec@1 60.62 #1 of 4 Archive leaderboard report
Image Classification JFT-300M V-MoE-H/14 (Last-5) prec@1 60.12 #2 of 4 Archive leaderboard report
Image Classification JFT-300M V-MoE-L/16 (Every-2) prec@1 57.65 #3 of 4 Archive leaderboard report
Image Classification JFT-300M VIT-H/14 prec@1 56.68 #4 of 4 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

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformerVision Transformer

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