AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

DDP-WM: Disentangled Dynamics Prediction for Efficient World Models

Shicheng Yin, Kaixuan Yin, Weixing Chen, Yang Liu, Guanbin Li, Liang Lin

ICML 2026 · Architectures and efficiency › Attention and transformers · approach: Transformer neural operators for PDEs · 0 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
add Low-Rank Correction Module (ablation)DDP-WM · Push-TMPC success rate (%): 62 -> 70Improved
add Low-Rank Correction Module (with MPC mask) (ablation)DDP-WM · Push-TMPC success rate (%): 70 -> 98Improved
add sparse MPC cost mask (ablation)DDP-WM · Push-TMPC success rate (%): 90 -> 98Improved
add sparse MPC cost mask (without LRM) (ablation)Naive Sparse · Push-TMPC success rate (%): 62 -> 70Improved
add LRM to naive sparse model, open-loop (ablation)DDP-WM · Push-T 5-step open-looppixel error: 427 -> 361Improved
add LRM to naive sparse model, open-loop (ablation)DDP-WM · Wall 5-step open-looppixel error: 15 -> 9Improved
add LRM to naive sparse model, open-loop (ablation)DDP-WM · PointMaze 5-step open-looppixel error: 41 -> 36Improved
add high-precision 2x2 sub-region localization (no LRM) (ablation)DDP-WM · Push-T 5-step open-looppixel error: 788 -> 427Improved
add high-precision localization (ablation)DDP-WM · Push-T mask qualityIoU: 0.3446 -> 0.8935Improved
sparse disentangled prediction vs dense DINO-WMDDP-WM · Push-TMPC success rate (%): 90 -> 98Improved
sparse disentangled prediction vs dense DINO-WMDDP-WM · Push-TMPC decision time (s): 120 -> 16Improved
sparse disentangled prediction vs dense DINO-WMDDP-WM · Push-TFLOPs per step (G): 23 -> 2.5Improved

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