Papers › Linear Attention with Global Context: A Multipole Attention Mechanism for Vision and Physics

Linear Attention with Global Context: A Multipole Attention Mechanism for Vision and Physics

3 Jul 2025arXiv:2507.02748archive 2025-07-28

Alex Colagrande, Paul Caillon, Eva Feillet, Alexandre Allauzen

Transformers have become the de facto standard for a wide range of tasks, from image classification to physics simulations. Despite their impressive performance, the quadratic complexity of standard Transformers in both memory and time with respect to the input length makes them impractical for processing high-resolution inputs. Therefore, several variants have been proposed, the most successful relying on patchification, downsampling, or coarsening techniques, often at the cost of losing the finest-scale details. In this work, we take a different approach. Inspired by state-of-the-art techniques in n-body numerical simulations, we cast attention as an interaction problem between grid points. We introduce the Multipole Attention Neural Operator (MANO), which computes attention in a distance-based multiscale fashion. MANO maintains, in each attention head, a global receptive field and achieves linear time and memory complexity with respect to the number of grid points. Empirical results on image classification and Darcy flows demonstrate that MANO rivals state-of-the-art models such as ViT and Swin Transformer, while reducing runtime and peak memory usage by orders of magnitude. We open source our code for reproducibility at https://github.com/AlexColagrande/MANO.

PaperPDFConference PDFCode

Code

AlexColagrande/MANO officialmentioned in papermentioned on GitHubpytorch 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

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Image Classificationimage-classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Fine-Grained Image Classification Oxford-IIIT Pet Dataset MANO-tiny Accuracy 88.31 #13 of 15 Archive leaderboard report
Fine-Grained Image Classification Stanford Cars MANO-tiny Accuracy 65.68 #83 of 83 Archive leaderboard report
Image Classification CIFAR-100 MANO-tiny Percentage correct 85.08 #69 of 211 Archive leaderboard report
Image Classification Flowers-102 MANO-tiny Accuracy 89.00 #49 of 52 Archive leaderboard report
Image Classification Food-101 MANO-tiny Accuracy (%) 82.48 #9 of 11 Archive leaderboard report
Image Classification Tiny ImageNet Classification MANO-tiny Validation Acc 87.52 #8 of 23 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.

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