AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Massive Memorization with Hundreds of Trillions of Parameters for Sequential Transducer Generative Recommenders

Zhimin Chen, Chenyu Zhao, Ka Chun Mo, Yunjiang Jiang, Jane H. Lee, Khushhall Chandra Mahajan, Ning Jiang, Kai Ren, Charlie Li, Wen-Yun Yang

ICLR 2026 · Architectures and efficiency › Attention and transformers · approach: State space and recurrent sequence models

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
Add generative reconstruction loss (ablation)VISTA · Industrial production eval NEEval NE: C-Task -0.29% -> -0.40%; E1-Task -1.29% vs -1.19%Mixed
Add generative reconstruction loss (ablation)VISTA · Amazon-ElectronicsAUC/NE: +0.22% AUC, 1.11% NEImproved
Use 256-dim instead of 128-dim embeddings (ablation)VISTA · Industrial eval NE C-TaskEval NE: -0.29% -> -0.40%Improved
Use 128 seeds instead of 64 (ablation)VISTA · Industrial eval NEEval NE: C-Task -0.37% -> -0.40%; E2 -3.01% vs -2.98%Mixed
Increase number of seed embeddings (ablation)VISTA · Minimal ProductionAUC/NEImproved
Use quasi-linear attention instead of softmax attention (ablation)VISTA · Minimal Production training timeepoch timeImproved
Use quasi-linear attention (longer seq, more layers) (ablation)VISTA · Industrial dataEval NE: -0.13% eval NE, +5% QPSImproved
Use VISTA instead of HSTU baselinesHSTU · Amazon-ElectronicsAUC: 0.884 -> 0.886Improved
Use VISTA instead of HSTU baselinesHSTU · Minimal ProductionAUC: 0.668 vs 0.627No clear effect
Use VISTA instead of HSTUHSTU · KuaiRand-1KNE: 0.863 -> 0.863No clear effect
Add VISTA to HSTU production modelHSTU · Online A/B 15 daysC-Task: +0.5%, 94% GPU reductionImproved
Causal vs full self attention for user history (ablation)VISTA · summarizationqualityNo clear effect

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