AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Structured Expert Routing with Multi-View Task Priors for Offline Meta-Reinforcement Learning

Yisen Zhao, Peixi Peng, Xinyu Hu, Cong Li, Zhan Su, Zhuojian Li

ICML 2026 · Reinforcement learning › Visual and representation RL · approach: Task structure and curricula for meta-RL

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 dynamics prior (ablation)Prompt-DT · Cheetah-Veltest return: -135.89 -> -112.51Improved
add dynamics prior (ablation)Prompt-DT · Hopper-Paramtest return: 309.90 -> 340.87Improved
add semantic prior on top of dynamics (ablation)Prompt-DT + Dyn · Hopper-Paramtest return: 340.87 -> 352.19Improved
add dynamics prior to MoE routing (ablation)MoE (M3DT) · Cheetah-Veltest return: -107.19 -> -106.49No clear effect
add semantic prior on top of dynamics (ablation)MoE + Dyn · Hopper-Paramtest return: 348.05 -> 364.59Improved
add trajectory prior on top of dynamics and semantic (ablation)MoE + Dyn + Sem · Cheetah-Veltest return: -102.88 -> -92.64Improved
combine Content and Global routing scores instead of Content only (ablation)TGR · Cheetah-Veltest return: -110.48 -> -92.64Improved
combine Content and Global routing scores instead of Global only (ablation)TGR · Hopper-Paramtest return: 362.45 -> 370.52Improved
add semantic input to routing (vs w/o Sem) (ablation)TGR · Cheetah-Veltest return: -105.75 -> -92.64Improved
add dynamics input to routing (vs w/o Dyn) (ablation)TGR · Hopper-Paramtest return: 360.18 -> 370.52Improved
add dynamics and trajectory inputs (vs w/o Dyn&Traj) (ablation)TGR · Cheetah-Veltest return: -120.18 -> -92.64Improved
use TGR instead of scaling Meta-DT to ~34M paramsMeta-DT (Large) · Hopper-Paramtest return: 350.47 -> 370.52Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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