AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

MoM: Linear Sequence Modeling with Mixture-of-Memories

Jusen Du, Weigao Sun, Disen Lan, Jiaxi Hu, Tao Zhang, Yu Cheng

ICLR 2026 · Architectures and efficiency › Attention and transformers · approach: State space and recurrent sequence models · 41 citations (OpenAlex)

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
use mixture of memories instead of single memory in Gated DeltaNetGated DeltaNet · Recall-intensive avg (380M)average score: 24.78 -> 28.16Improved
use mixture of memories instead of single memory in Gated DeltaNetGated DeltaNet · Recall-intensive avg (1.3B)average score: 32.30 -> 36.04Improved
use MoM instead of Transformer++ on recall tasksTransformer++ · Recall-intensive avg (1.3B)average score: 37.31 -> 36.04Hurt
use MoM instead of Transformer++ at 380MTransformer++ · Recall-intensive avg (380M)average score: 31.70 -> 28.16Hurt
separate memories instead of expanding a single memory (ablation)Gated DeltaNet · Recall-intensive avg (380M)average score: 26.32 -> 28.16Improved
separate memories instead of expanding a single memory (ablation)GLA · Recall-intensive avg (380M)average score: 22.87 -> 23.53Improved
equalize activated parameters (400M) for MoM vs Gated DeltaNetGated DeltaNet · Recall-intensive avg (380M)average score: 24.78 -> 26.51Improved
equalize activated parameters on commonsenseGated DeltaNet · Commonsense avg (400M activated)average accuracy: 41.44 -> 41.50No clear effect
MoM on commonsense at 380M vs Gated DeltaNetGated DeltaNet · Commonsense avg (380M)average accuracy: 41.44 -> 41.97Improved
add 3 Transformer layers to MoM (hybrid) (ablation)MoM · Recall-intensive avg (380M)average score: 28.16 -> 37.80Improved
add 3 Transformer layers to MoM (hybrid) (ablation)MoM · Commonsense avg (380M)average accuracy: 41.97 -> 42.38Improved
increase number of memories from 1 to 8 at activation ratio 0.5 (ablation)MoM · recall and commonsense benchmarksaverageImproved

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