AI Paper Atlas

Research map / Reinforcement learning / Offline RL

MAGE: Multi-scale Autoregressive Generation for Offline Reinforcement Learning

Chenxing Lin, Xinhui Gao, Haipeng Zhang, Xinran Li, Haitao Wang, Songzhu Mei, Chenglu Wen, Weiquan Liu, Siqi Shen, Cheng Wang

ICLR 2026 · Reinforcement learning › Offline RL · approach: Diffusion models for offline decision making · 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 multi-scale modeling: K=1 to K=8 (ablation)MAGE · Adroit Pen-Expertnormalized score: 123.5 -> 147.8Improved
increase temporal scales from K=8 to K=10 (ablation)MAGE · Adroit Door-Clonednormalized score: 20.5 -> 17.0Hurt
increase temporal scales from K=8 to K=10 (ablation)MAGE · Adroit Pen-Expertnormalized score: 147.8 -> 149.9No clear effect
model returns and states (R,S) instead of states only (ablation)MAGE · Adroit Pen-Expertnormalized score: 127.7 -> 147.8Improved
model returns and states instead of returns, states and actions (ablation)MAGE · Adroit Door-Clonednormalized score: 17.2 -> 20.5Improved
model returns and states instead of actions with CQL (ablation)MAGE · Adroit Door-Clonednormalized score: 4.9 -> 20.5Improved
add RTG conditioning in the autoencoder (ablation)MAGE · Adroit Pen-Expertnormalized score: 140.3 -> 147.8Improved
add RTG conditioning in the autoencoder (ablation)MAGE · Adroit Door-Clonednormalized score: 12.3 -> 20.5Improved
add RTG conditioning in transformer for finer scales (ablation)MAGE · Adroit Door-Clonednormalized score: 16.3 -> 20.5Improved
add RTG in conditioning loss (ablation)MAGE · Adroit Pen-Expertnormalized score: 139.9 -> 147.8Improved
replace hierarchical diffusion with MAGEHD · Adroit (mean all settings)normalized score: 40.3 -> 66.9Improved
replace HD with MAGEHD · Franka Kitchenaverage score: 72.5 -> 88.8Improved

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