AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

MIMOMamba: From Scalar Duality to Matrix-Valued Attention

Yanbo Li, Richard Cornelius Suwandi, Feng Yin, SUN Yiyong, Wei Huang, Wenqiang Pu

ICML 2026 · Architectures and efficiency › Attention and transformers · approach: Mamba architectures

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Transformer with MIMOMambaSSP forecaster · SSP predictionRMSE mean: 0.748 -> 0.688Improved
replace Transformer with MIMOMambaSSP forecaster · SSP predictionParams: 100,801 -> 35,593Improved
replace LSTM with MIMOMambaSSP forecaster · SSP predictionRMSE max: 2.993 -> 1.662Improved
replace Transformer with MIMOMamba on thermocline regionSSP forecaster · SSP thermocline (~30 m)RMSE max: 14% reductionImproved
compare MIMOMamba against Mamba-2 and Mamba-3MIMOMamba · SSP prediction (extended)RMSE mean: 0.717 (Mamba-2), 0.715 (Mamba-3) -> 0.687Improved
use MIMOMamba on SCP1 versus SISO baselinesMIMOMamba · UEA SCP1accuracy: 85.5±2.9%Improved
vary polynomial degree p from 2 to 8 (ablation)MIMOMamba · SSP predictionRMSE: p = 4 sweet spotMixed
use matrix polynomial sharing instead of independent dense A, B, CMIMOMamba · core dynamicsparameter count: 3D^2 -> 1D^2Improved
use FFT polynomial multiplication at degree 4MIMOMamba · current configurationspeedNo clear effect
use MIMOMamba-Fast instead of Mamba-2 in pure PyTorchMIMOMamba-Fast · Training throughputthroughput: 1.5-1.6x fasterImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy