AI Paper Atlas

Research map / Architectures and efficiency / Mixture of experts

One Model for All Tasks: Leveraging Efficient World Models in Multi-Task Planning

Yuan Pu, Yazhe Niu, Jia Tang, Junyu Xiong, Shuai Hu, Hongsheng Li

ICLR 2026 · Architectures and efficiency › Mixture of experts · approach: Sparse MoE as regularization and attention · 5 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
replace dense FFN with sparse MoE backbone in UniZero (ablation)UniZero · Atari8 multitaskreturn / plasticity metricsImproved
add explicit learnable task embedding (ablation)UniZero · Atari8 multitaskreturnNo clear effect
replace ResNet encoder with ViT (ablation)UniZero · Atari8 multitaskreturnNo clear effect
replace LayerNorm with SimNorm (ablation)UniZero · Atari8 multitaskreturnMixed
add MoCo gradient correction (ablation)UniZero · multitask suitereturn and training time: approximately 40%Hurt
use frozen DINOv2 encoder instead of from-scratch ViT (ablation)ScaleZero · AtarireturnNo clear effect
train multitask ScaleZero instead of multitask UniZeroUniZero (MT) · Atari 100k (26 games)human-normalized mean: 0.31 -> 0.39Improved
multitask ScaleZero vs 26 single-task UniZeroUniZero (ST) · Atari 100k (26 games)human-normalized median: 0.21 -> 0.16Hurt
multitask ScaleZero vs 18 single-task UniZeroUniZero (ST) · DMControl (18 tasks)mean return: 787.2 -> 769.7Hurt
multitask ScaleZero vs 18 single-task UniZeroUniZero (ST) · DMControl (18 tasks)median return: 875.1 -> 887.3Improved
multitask ScaleZero vs single-task UniZeroUniZero (ST) · Jericho Detectiveaverage return: 295 -> 280Hurt
add Dynamic Parameter Scaling (LoRA stages)ScaleZero · DMControl (18 tasks)environment interactions: 71.5%Improved

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