AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Composition of Memory Experts for Diffusion World Models

Sebastian Stapf, Pablo Acuaviva, Aram Davtyan, Paolo Favaro

ICLR 2026 · Architectures and efficiency › Attention and transformers · approach: State space and recurrent sequence models · 1 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 contrastive product of experts instead of plain PoE when combining all experts (ablation)CoME (all experts) · Memory MazeLPIPS: 0.192 -> 0.097Improved
add contrastive weighting to STM+LTM (ablation)CoME · Memory MazeLPIPS: 0.170 -> 0.114Improved
add contrastive baseline to the base model alone (ablation)Base DiT · Memory MazeLPIPS: 0.203 -> 0.200No clear effect
add long-term memory via LoRA test-time finetuning (ablation)base DiT · Memory MazeLPIPS: 0.203 -> 0.171Improved
combine STM and LTM experts (ablation)CoME · Memory MazeLPIPS: 0.171 -> 0.114Improved
increase LTM context frames from 50 to 450 (ablation)LTM (LoRA rank 256) · Memory MazeLPIPS: 0.162 -> 0.118Improved
use LoRA instead of full finetuning for LTM at low context (ablation)LTM · Memory Maze (50 frames)LPIPS: 0.221 (full) -> 0.162 (rank 256)Improved
add STM or LTM memory to NWM plannerNWM · RECON planningATE: 1.13 -> 0.96 (CoME)Improved
add memory experts to Oasis-style Minecraft world modelbase diffusion WM · Minecraft MarshPSNR: 16.19 -> 17.78Improved
add Langevin correction steps during sampling (ablation)CoME · Memory Mazegeneration qualityNo clear effect
increase LoRA rank for LTM memorization (compute cost) (ablation)LTM · sampling overheadsampling time overhead: r=64 about 4x vs r=16 about 2xHurt

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